Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

added_tokens.json

@@ -0,0 +1,28 @@
+{
+  "</think>": 151668,
+  "</tool_call>": 151658,
+  "</tool_response>": 151666,
+  "<think>": 151667,
+  "<tool_call>": 151657,
+  "<tool_response>": 151665,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

config.json

@@ -0,0 +1,38 @@
+{
+  "architectures": [
+    "Qwen3MoeForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "decoder_sparse_step": 1,
+  "eos_token_id": 151645,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "initializer_range": 0.02,
+  "intermediate_size": 6144,
+  "max_position_embeddings": 40960,
+  "max_window_layers": 48,
+  "mlp_only_layers": [],
+  "model_type": "qwen3_moe",
+  "moe_intermediate_size": 768,
+  "norm_topk_prob": true,
+  "num_attention_heads": 32,
+  "num_experts": 128,
+  "num_experts_per_tok": 8,
+  "num_hidden_layers": 48,
+  "num_key_value_heads": 4,
+  "output_router_logits": false,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "router_aux_loss_coef": 0.001,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.51.3",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

generation_config.json

@@ -0,0 +1,13 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "temperature": 0.6,
+  "top_k": 20,
+  "top_p": 0.95,
+  "transformers_version": "4.51.3"
+}

model-00001-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f9700dc7fc947a2f1da8674f29b82e37c5d73767574439b3f4777f6be3de0c20
+size 4997184968

model-00002-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3983de03e7ecdbafb3648173edd54c27042ff35d9fc9740172744a3c7419208
+size 4997741608

model-00003-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c565d191abf551e54b03b7b55ba8ac9478c50d32f51e2d23097847a61e047a48
+size 4997742208

model-00004-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b3fac05816e8367132690cd54d952bdee749fec505a39ef1e5c661f18fab16b
+size 4997743184

model-00005-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3288fa2e75ee7aa4ad7139f35ae0a47ad6f470969d5653e6a3e5feb47c164e4c
+size 4997743184

model-00006-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c905522435e9c9309ef82decc452eaa7070c36eda05f04701edfae0d93c9c724
+size 4997743184

model-00007-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5c85a0da305f5761d3b73582fe2eaf71b7aeb8741134f2ec8ac6dc133e608a9
+size 4997743184

model-00008-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6408221b17e17ae78aed4a6e5083c5bcca806485b176ef7cd53c57911d79591e
+size 4997743184

model-00009-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2654051549ef49d8fa9c481130f48234f3ce736f3fb96d4c2d01942495651f25
+size 4997743184

model-00010-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:940297f79ac5d6146021950c1913588a53041e2c7941a848e5c8c25d5293a3ac
+size 4997743184

model-00011-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aecb538e5072ea5ca0ab77753d2f88a0079a3593a8b0bf9386bf277ab5b3af70
+size 4997743184

model-00012-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8953c91239fa8ec179c0e8dc0dd0f271119c353696000579b22fee19096580dd
+size 4997743184

model-00013-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c8f4fbe63dd205ac08157aab6113e2130cce876dbe4ec03b224a953ab284e30b
+size 1094220288

special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "eos_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aeb13307a71acd8fe81861d94ad54ab689df773318809eed3cbe794b4492dae4
+size 11422654

tokenizer_config.json

@@ -0,0 +1,241 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151665": {
+      "content": "<tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151666": {
+      "content": "</tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151667": {
+      "content": "<think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151668": {
+      "content": "</think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0].role == 'system' %}\n        {{- messages[0].content + '\\n\\n' }}\n    {%- endif %}\n    {{- \"# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0].role == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0].content + '<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}\n{%- for message in messages[::-1] %}\n    {%- set index = (messages|length - 1) - loop.index0 %}\n    {%- if ns.multi_step_tool and message.role == \"user\" and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}\n        {%- set ns.multi_step_tool = false %}\n        {%- set ns.last_query_index = index %}\n    {%- endif %}\n{%- endfor %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {%- set content = message.content %}\n        {%- set reasoning_content = '' %}\n        {%- if message.reasoning_content is defined and message.reasoning_content is not none %}\n            {%- set reasoning_content = message.reasoning_content %}\n        {%- else %}\n            {%- if '</think>' in message.content %}\n                {%- set content = message.content.split('</think>')[-1].lstrip('\\n') %}\n                {%- set reasoning_content = message.content.split('</think>')[0].rstrip('\\n').split('<think>')[-1].lstrip('\\n') %}\n            {%- endif %}\n        {%- endif %}\n        {%- if loop.index0 > ns.last_query_index %}\n            {%- if loop.last or (not loop.last and reasoning_content) %}\n                {{- '<|im_start|>' + message.role + '\\n<think>\\n' + reasoning_content.strip('\\n') + '\\n</think>\\n\\n' + content.lstrip('\\n') }}\n            {%- else %}\n                {{- '<|im_start|>' + message.role + '\\n' + content }}\n            {%- endif %}\n        {%- else %}\n            {{- '<|im_start|>' + message.role + '\\n' + content }}\n        {%- endif %}\n        {%- if message.tool_calls %}\n            {%- for tool_call in message.tool_calls %}\n                {%- if (loop.first and content) or (not loop.first) %}\n                    {{- '\\n' }}\n                {%- endif %}\n                {%- if tool_call.function %}\n                    {%- set tool_call = tool_call.function %}\n                {%- endif %}\n                {{- '<tool_call>\\n{\"name\": \"' }}\n                {{- tool_call.name }}\n                {{- '\", \"arguments\": ' }}\n                {%- if tool_call.arguments is string %}\n                    {{- tool_call.arguments }}\n                {%- else %}\n                    {{- tool_call.arguments | tojson }}\n                {%- endif %}\n                {{- '}\\n</tool_call>' }}\n            {%- endfor %}\n        {%- endif %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if loop.first or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n    {%- if enable_thinking is defined and enable_thinking is false %}\n        {{- '<think>\\n\\n</think>\\n\\n' }}\n    {%- endif %}\n{%- endif %}",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

trainer_state.json

@@ -0,0 +1,3310 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.984025559105431,
+  "eval_steps": 500,
+  "global_step": 468,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.006389776357827476,
+      "grad_norm": 31.704153653628946,
+      "learning_rate": 0.0,
+      "loss": 2.5768,
+      "step": 1
+    },
+    {
+      "epoch": 0.012779552715654952,
+      "grad_norm": 31.58442561735789,
+      "learning_rate": 2.1276595744680852e-07,
+      "loss": 2.7208,
+      "step": 2
+    },
+    {
+      "epoch": 0.019169329073482427,
+      "grad_norm": 31.41195342659084,
+      "learning_rate": 4.2553191489361704e-07,
+      "loss": 2.5583,
+      "step": 3
+    },
+    {
+      "epoch": 0.025559105431309903,
+      "grad_norm": 33.9099531804883,
+      "learning_rate": 6.382978723404255e-07,
+      "loss": 2.6238,
+      "step": 4
+    },
+    {
+      "epoch": 0.03194888178913738,
+      "grad_norm": 27.556054406511105,
+      "learning_rate": 8.510638297872341e-07,
+      "loss": 2.4049,
+      "step": 5
+    },
+    {
+      "epoch": 0.038338658146964855,
+      "grad_norm": 31.768418102023194,
+      "learning_rate": 1.0638297872340427e-06,
+      "loss": 2.6021,
+      "step": 6
+    },
+    {
+      "epoch": 0.04472843450479233,
+      "grad_norm": 32.283847840303004,
+      "learning_rate": 1.276595744680851e-06,
+      "loss": 2.743,
+      "step": 7
+    },
+    {
+      "epoch": 0.051118210862619806,
+      "grad_norm": 29.455083723950487,
+      "learning_rate": 1.4893617021276596e-06,
+      "loss": 2.4725,
+      "step": 8
+    },
+    {
+      "epoch": 0.05750798722044728,
+      "grad_norm": 27.236601491186654,
+      "learning_rate": 1.7021276595744682e-06,
+      "loss": 2.3754,
+      "step": 9
+    },
+    {
+      "epoch": 0.06389776357827476,
+      "grad_norm": 29.533151207286714,
+      "learning_rate": 1.9148936170212767e-06,
+      "loss": 2.3779,
+      "step": 10
+    },
+    {
+      "epoch": 0.07028753993610223,
+      "grad_norm": 23.426628166758984,
+      "learning_rate": 2.1276595744680853e-06,
+      "loss": 2.2612,
+      "step": 11
+    },
+    {
+      "epoch": 0.07667731629392971,
+      "grad_norm": 24.604043292573415,
+      "learning_rate": 2.340425531914894e-06,
+      "loss": 2.0661,
+      "step": 12
+    },
+    {
+      "epoch": 0.08306709265175719,
+      "grad_norm": 24.655783466838816,
+      "learning_rate": 2.553191489361702e-06,
+      "loss": 2.2323,
+      "step": 13
+    },
+    {
+      "epoch": 0.08945686900958466,
+      "grad_norm": 15.84080580636176,
+      "learning_rate": 2.765957446808511e-06,
+      "loss": 1.7864,
+      "step": 14
+    },
+    {
+      "epoch": 0.09584664536741214,
+      "grad_norm": 12.05049966753011,
+      "learning_rate": 2.978723404255319e-06,
+      "loss": 1.6959,
+      "step": 15
+    },
+    {
+      "epoch": 0.10223642172523961,
+      "grad_norm": 12.020085870345559,
+      "learning_rate": 3.191489361702128e-06,
+      "loss": 1.5898,
+      "step": 16
+    },
+    {
+      "epoch": 0.10862619808306709,
+      "grad_norm": 12.3340962254883,
+      "learning_rate": 3.4042553191489363e-06,
+      "loss": 1.6419,
+      "step": 17
+    },
+    {
+      "epoch": 0.11501597444089456,
+      "grad_norm": 8.611078574621429,
+      "learning_rate": 3.6170212765957453e-06,
+      "loss": 1.3988,
+      "step": 18
+    },
+    {
+      "epoch": 0.12140575079872204,
+      "grad_norm": 4.680603958774463,
+      "learning_rate": 3.8297872340425535e-06,
+      "loss": 1.2925,
+      "step": 19
+    },
+    {
+      "epoch": 0.12779552715654952,
+      "grad_norm": 6.743115476232132,
+      "learning_rate": 4.042553191489362e-06,
+      "loss": 1.3818,
+      "step": 20
+    },
+    {
+      "epoch": 0.134185303514377,
+      "grad_norm": 4.157596387709073,
+      "learning_rate": 4.255319148936171e-06,
+      "loss": 1.3793,
+      "step": 21
+    },
+    {
+      "epoch": 0.14057507987220447,
+      "grad_norm": 2.9340588329022443,
+      "learning_rate": 4.468085106382979e-06,
+      "loss": 1.2788,
+      "step": 22
+    },
+    {
+      "epoch": 0.14696485623003194,
+      "grad_norm": 2.818255229706057,
+      "learning_rate": 4.680851063829788e-06,
+      "loss": 1.1308,
+      "step": 23
+    },
+    {
+      "epoch": 0.15335463258785942,
+      "grad_norm": 3.117187687433535,
+      "learning_rate": 4.893617021276596e-06,
+      "loss": 1.3701,
+      "step": 24
+    },
+    {
+      "epoch": 0.1597444089456869,
+      "grad_norm": 2.8101978209724914,
+      "learning_rate": 5.106382978723404e-06,
+      "loss": 1.2065,
+      "step": 25
+    },
+    {
+      "epoch": 0.16613418530351437,
+      "grad_norm": 2.559531586573765,
+      "learning_rate": 5.319148936170213e-06,
+      "loss": 1.2604,
+      "step": 26
+    },
+    {
+      "epoch": 0.17252396166134185,
+      "grad_norm": 2.5602544421302627,
+      "learning_rate": 5.531914893617022e-06,
+      "loss": 1.1226,
+      "step": 27
+    },
+    {
+      "epoch": 0.17891373801916932,
+      "grad_norm": 2.78567947341407,
+      "learning_rate": 5.744680851063831e-06,
+      "loss": 1.1317,
+      "step": 28
+    },
+    {
+      "epoch": 0.1853035143769968,
+      "grad_norm": 2.32140520760061,
+      "learning_rate": 5.957446808510638e-06,
+      "loss": 1.2707,
+      "step": 29
+    },
+    {
+      "epoch": 0.19169329073482427,
+      "grad_norm": 2.637245330226456,
+      "learning_rate": 6.170212765957447e-06,
+      "loss": 1.1876,
+      "step": 30
+    },
+    {
+      "epoch": 0.19808306709265175,
+      "grad_norm": 2.655115239155347,
+      "learning_rate": 6.382978723404256e-06,
+      "loss": 1.1477,
+      "step": 31
+    },
+    {
+      "epoch": 0.20447284345047922,
+      "grad_norm": 2.1291613777910037,
+      "learning_rate": 6.595744680851064e-06,
+      "loss": 1.2044,
+      "step": 32
+    },
+    {
+      "epoch": 0.2108626198083067,
+      "grad_norm": 2.2821733814301273,
+      "learning_rate": 6.808510638297873e-06,
+      "loss": 1.2596,
+      "step": 33
+    },
+    {
+      "epoch": 0.21725239616613418,
+      "grad_norm": 2.1391452019594435,
+      "learning_rate": 7.021276595744682e-06,
+      "loss": 1.1544,
+      "step": 34
+    },
+    {
+      "epoch": 0.22364217252396165,
+      "grad_norm": 2.634505180459112,
+      "learning_rate": 7.234042553191491e-06,
+      "loss": 1.1743,
+      "step": 35
+    },
+    {
+      "epoch": 0.23003194888178913,
+      "grad_norm": 2.4479669174854695,
+      "learning_rate": 7.446808510638298e-06,
+      "loss": 1.2329,
+      "step": 36
+    },
+    {
+      "epoch": 0.2364217252396166,
+      "grad_norm": 2.4211752635431045,
+      "learning_rate": 7.659574468085107e-06,
+      "loss": 1.2652,
+      "step": 37
+    },
+    {
+      "epoch": 0.24281150159744408,
+      "grad_norm": 2.4226771286938864,
+      "learning_rate": 7.872340425531916e-06,
+      "loss": 1.2498,
+      "step": 38
+    },
+    {
+      "epoch": 0.24920127795527156,
+      "grad_norm": 2.243174790855972,
+      "learning_rate": 8.085106382978723e-06,
+      "loss": 1.0903,
+      "step": 39
+    },
+    {
+      "epoch": 0.25559105431309903,
+      "grad_norm": 2.194169145672455,
+      "learning_rate": 8.297872340425532e-06,
+      "loss": 1.2425,
+      "step": 40
+    },
+    {
+      "epoch": 0.26198083067092653,
+      "grad_norm": 2.3929823585508028,
+      "learning_rate": 8.510638297872341e-06,
+      "loss": 1.2134,
+      "step": 41
+    },
+    {
+      "epoch": 0.268370607028754,
+      "grad_norm": 2.192126133315745,
+      "learning_rate": 8.72340425531915e-06,
+      "loss": 1.2291,
+      "step": 42
+    },
+    {
+      "epoch": 0.2747603833865815,
+      "grad_norm": 2.076981829205724,
+      "learning_rate": 8.936170212765958e-06,
+      "loss": 1.0894,
+      "step": 43
+    },
+    {
+      "epoch": 0.28115015974440893,
+      "grad_norm": 2.159545856122453,
+      "learning_rate": 9.148936170212767e-06,
+      "loss": 1.1815,
+      "step": 44
+    },
+    {
+      "epoch": 0.28753993610223644,
+      "grad_norm": 1.9857699802597364,
+      "learning_rate": 9.361702127659576e-06,
+      "loss": 1.0501,
+      "step": 45
+    },
+    {
+      "epoch": 0.2939297124600639,
+      "grad_norm": 2.0870694200309265,
+      "learning_rate": 9.574468085106385e-06,
+      "loss": 1.1938,
+      "step": 46
+    },
+    {
+      "epoch": 0.3003194888178914,
+      "grad_norm": 2.3350367723863084,
+      "learning_rate": 9.787234042553192e-06,
+      "loss": 1.0923,
+      "step": 47
+    },
+    {
+      "epoch": 0.30670926517571884,
+      "grad_norm": 2.455931252583481,
+      "learning_rate": 1e-05,
+      "loss": 1.1516,
+      "step": 48
+    },
+    {
+      "epoch": 0.31309904153354634,
+      "grad_norm": 2.1368969102758424,
+      "learning_rate": 9.999860789001947e-06,
+      "loss": 0.9968,
+      "step": 49
+    },
+    {
+      "epoch": 0.3194888178913738,
+      "grad_norm": 2.3688426509121596,
+      "learning_rate": 9.999443163759669e-06,
+      "loss": 1.1361,
+      "step": 50
+    },
+    {
+      "epoch": 0.3258785942492013,
+      "grad_norm": 2.3264607493113325,
+      "learning_rate": 9.998747147528375e-06,
+      "loss": 1.1678,
+      "step": 51
+    },
+    {
+      "epoch": 0.33226837060702874,
+      "grad_norm": 2.0130202409916786,
+      "learning_rate": 9.997772779065312e-06,
+      "loss": 1.1882,
+      "step": 52
+    },
+    {
+      "epoch": 0.33865814696485624,
+      "grad_norm": 2.39774596578351,
+      "learning_rate": 9.996520112627602e-06,
+      "loss": 1.2381,
+      "step": 53
+    },
+    {
+      "epoch": 0.3450479233226837,
+      "grad_norm": 2.3378114420187295,
+      "learning_rate": 9.994989217969224e-06,
+      "loss": 1.292,
+      "step": 54
+    },
+    {
+      "epoch": 0.3514376996805112,
+      "grad_norm": 2.4669226231480037,
+      "learning_rate": 9.993180180337126e-06,
+      "loss": 1.2083,
+      "step": 55
+    },
+    {
+      "epoch": 0.35782747603833864,
+      "grad_norm": 2.5723057747722446,
+      "learning_rate": 9.991093100466482e-06,
+      "loss": 1.2001,
+      "step": 56
+    },
+    {
+      "epoch": 0.36421725239616615,
+      "grad_norm": 2.1243722512989254,
+      "learning_rate": 9.988728094575082e-06,
+      "loss": 1.213,
+      "step": 57
+    },
+    {
+      "epoch": 0.3706070287539936,
+      "grad_norm": 2.4033633557830343,
+      "learning_rate": 9.986085294356858e-06,
+      "loss": 1.1875,
+      "step": 58
+    },
+    {
+      "epoch": 0.3769968051118211,
+      "grad_norm": 2.3474389927950385,
+      "learning_rate": 9.983164846974549e-06,
+      "loss": 1.2322,
+      "step": 59
+    },
+    {
+      "epoch": 0.38338658146964855,
+      "grad_norm": 2.452909045917012,
+      "learning_rate": 9.979966915051517e-06,
+      "loss": 1.1646,
+      "step": 60
+    },
+    {
+      "epoch": 0.38977635782747605,
+      "grad_norm": 2.4723875025133837,
+      "learning_rate": 9.976491676662679e-06,
+      "loss": 1.3015,
+      "step": 61
+    },
+    {
+      "epoch": 0.3961661341853035,
+      "grad_norm": 2.2634946397130347,
+      "learning_rate": 9.972739325324596e-06,
+      "loss": 1.2803,
+      "step": 62
+    },
+    {
+      "epoch": 0.402555910543131,
+      "grad_norm": 2.403155802895676,
+      "learning_rate": 9.968710069984699e-06,
+      "loss": 1.2175,
+      "step": 63
+    },
+    {
+      "epoch": 0.40894568690095845,
+      "grad_norm": 2.1295330822670127,
+      "learning_rate": 9.964404135009649e-06,
+      "loss": 1.2292,
+      "step": 64
+    },
+    {
+      "epoch": 0.41533546325878595,
+      "grad_norm": 2.1414114768880808,
+      "learning_rate": 9.959821760172849e-06,
+      "loss": 1.2264,
+      "step": 65
+    },
+    {
+      "epoch": 0.4217252396166134,
+      "grad_norm": 2.2853864319073787,
+      "learning_rate": 9.95496320064109e-06,
+      "loss": 1.1838,
+      "step": 66
+    },
+    {
+      "epoch": 0.4281150159744409,
+      "grad_norm": 2.2370120573822874,
+      "learning_rate": 9.94982872696034e-06,
+      "loss": 1.291,
+      "step": 67
+    },
+    {
+      "epoch": 0.43450479233226835,
+      "grad_norm": 2.2059962759749063,
+      "learning_rate": 9.94441862504068e-06,
+      "loss": 1.165,
+      "step": 68
+    },
+    {
+      "epoch": 0.44089456869009586,
+      "grad_norm": 2.5409369250481957,
+      "learning_rate": 9.938733196140386e-06,
+      "loss": 1.2124,
+      "step": 69
+    },
+    {
+      "epoch": 0.4472843450479233,
+      "grad_norm": 2.0755145142571996,
+      "learning_rate": 9.932772756849152e-06,
+      "loss": 1.3153,
+      "step": 70
+    },
+    {
+      "epoch": 0.4536741214057508,
+      "grad_norm": 2.111481371806788,
+      "learning_rate": 9.926537639070457e-06,
+      "loss": 1.208,
+      "step": 71
+    },
+    {
+      "epoch": 0.46006389776357826,
+      "grad_norm": 2.309594198706378,
+      "learning_rate": 9.92002819000309e-06,
+      "loss": 1.1227,
+      "step": 72
+    },
+    {
+      "epoch": 0.46645367412140576,
+      "grad_norm": 1.812782826112667,
+      "learning_rate": 9.913244772121811e-06,
+      "loss": 1.2081,
+      "step": 73
+    },
+    {
+      "epoch": 0.4728434504792332,
+      "grad_norm": 2.265672430816926,
+      "learning_rate": 9.90618776315717e-06,
+      "loss": 1.3102,
+      "step": 74
+    },
+    {
+      "epoch": 0.4792332268370607,
+      "grad_norm": 2.141190910110161,
+      "learning_rate": 9.898857556074469e-06,
+      "loss": 1.151,
+      "step": 75
+    },
+    {
+      "epoch": 0.48562300319488816,
+      "grad_norm": 2.1944205784706474,
+      "learning_rate": 9.891254559051886e-06,
+      "loss": 1.236,
+      "step": 76
+    },
+    {
+      "epoch": 0.49201277955271566,
+      "grad_norm": 2.2423059756988923,
+      "learning_rate": 9.883379195457747e-06,
+      "loss": 1.1508,
+      "step": 77
+    },
+    {
+      "epoch": 0.4984025559105431,
+      "grad_norm": 2.3279349540321492,
+      "learning_rate": 9.875231903826936e-06,
+      "loss": 1.0754,
+      "step": 78
+    },
+    {
+      "epoch": 0.5047923322683706,
+      "grad_norm": 2.053366178339911,
+      "learning_rate": 9.8668131378365e-06,
+      "loss": 1.2044,
+      "step": 79
+    },
+    {
+      "epoch": 0.5111821086261981,
+      "grad_norm": 2.126844691744352,
+      "learning_rate": 9.858123366280358e-06,
+      "loss": 1.1222,
+      "step": 80
+    },
+    {
+      "epoch": 0.5175718849840255,
+      "grad_norm": 2.3371921606203365,
+      "learning_rate": 9.849163073043223e-06,
+      "loss": 1.2579,
+      "step": 81
+    },
+    {
+      "epoch": 0.5239616613418531,
+      "grad_norm": 1.9862725064766693,
+      "learning_rate": 9.83993275707364e-06,
+      "loss": 1.128,
+      "step": 82
+    },
+    {
+      "epoch": 0.5303514376996805,
+      "grad_norm": 2.0299267603970046,
+      "learning_rate": 9.830432932356207e-06,
+      "loss": 1.06,
+      "step": 83
+    },
+    {
+      "epoch": 0.536741214057508,
+      "grad_norm": 1.967123761472139,
+      "learning_rate": 9.820664127882958e-06,
+      "loss": 1.1022,
+      "step": 84
+    },
+    {
+      "epoch": 0.5431309904153354,
+      "grad_norm": 2.0479944136804407,
+      "learning_rate": 9.8106268876239e-06,
+      "loss": 1.3196,
+      "step": 85
+    },
+    {
+      "epoch": 0.549520766773163,
+      "grad_norm": 2.0899371871761714,
+      "learning_rate": 9.800321770496726e-06,
+      "loss": 1.2082,
+      "step": 86
+    },
+    {
+      "epoch": 0.5559105431309904,
+      "grad_norm": 1.9264654398492607,
+      "learning_rate": 9.789749350335693e-06,
+      "loss": 1.0647,
+      "step": 87
+    },
+    {
+      "epoch": 0.5623003194888179,
+      "grad_norm": 2.404003249982565,
+      "learning_rate": 9.778910215859666e-06,
+      "loss": 1.1614,
+      "step": 88
+    },
+    {
+      "epoch": 0.5686900958466453,
+      "grad_norm": 2.2535640052562123,
+      "learning_rate": 9.767804970639338e-06,
+      "loss": 1.29,
+      "step": 89
+    },
+    {
+      "epoch": 0.5750798722044729,
+      "grad_norm": 1.7435884422501948,
+      "learning_rate": 9.756434233063616e-06,
+      "loss": 1.1508,
+      "step": 90
+    },
+    {
+      "epoch": 0.5814696485623003,
+      "grad_norm": 2.1089410317587602,
+      "learning_rate": 9.744798636305189e-06,
+      "loss": 1.0881,
+      "step": 91
+    },
+    {
+      "epoch": 0.5878594249201278,
+      "grad_norm": 1.998448289395409,
+      "learning_rate": 9.732898828285273e-06,
+      "loss": 1.2313,
+      "step": 92
+    },
+    {
+      "epoch": 0.5942492012779552,
+      "grad_norm": 2.167555925222082,
+      "learning_rate": 9.72073547163753e-06,
+      "loss": 1.1593,
+      "step": 93
+    },
+    {
+      "epoch": 0.6006389776357828,
+      "grad_norm": 2.057046306061941,
+      "learning_rate": 9.708309243671167e-06,
+      "loss": 1.1188,
+      "step": 94
+    },
+    {
+      "epoch": 0.6070287539936102,
+      "grad_norm": 2.0592362759109073,
+      "learning_rate": 9.695620836333219e-06,
+      "loss": 1.2007,
+      "step": 95
+    },
+    {
+      "epoch": 0.6134185303514377,
+      "grad_norm": 2.212221040092753,
+      "learning_rate": 9.68267095617003e-06,
+      "loss": 1.2359,
+      "step": 96
+    },
+    {
+      "epoch": 0.6198083067092651,
+      "grad_norm": 2.0841357313666706,
+      "learning_rate": 9.669460324287899e-06,
+      "loss": 1.0926,
+      "step": 97
+    },
+    {
+      "epoch": 0.6261980830670927,
+      "grad_norm": 2.076701412548829,
+      "learning_rate": 9.655989676312918e-06,
+      "loss": 1.3485,
+      "step": 98
+    },
+    {
+      "epoch": 0.6325878594249201,
+      "grad_norm": 1.9517624724174427,
+      "learning_rate": 9.642259762350034e-06,
+      "loss": 1.149,
+      "step": 99
+    },
+    {
+      "epoch": 0.6389776357827476,
+      "grad_norm": 2.2261739182237834,
+      "learning_rate": 9.628271346941252e-06,
+      "loss": 1.1836,
+      "step": 100
+    },
+    {
+      "epoch": 0.645367412140575,
+      "grad_norm": 1.92940622703802,
+      "learning_rate": 9.614025209023084e-06,
+      "loss": 1.0385,
+      "step": 101
+    },
+    {
+      "epoch": 0.6517571884984026,
+      "grad_norm": 2.136053016131896,
+      "learning_rate": 9.59952214188316e-06,
+      "loss": 1.1348,
+      "step": 102
+    },
+    {
+      "epoch": 0.65814696485623,
+      "grad_norm": 2.3240551796672286,
+      "learning_rate": 9.58476295311606e-06,
+      "loss": 1.1955,
+      "step": 103
+    },
+    {
+      "epoch": 0.6645367412140575,
+      "grad_norm": 1.9948022400284593,
+      "learning_rate": 9.569748464578343e-06,
+      "loss": 1.3491,
+      "step": 104
+    },
+    {
+      "epoch": 0.670926517571885,
+      "grad_norm": 2.298108412072366,
+      "learning_rate": 9.554479512342785e-06,
+      "loss": 1.1793,
+      "step": 105
+    },
+    {
+      "epoch": 0.6773162939297125,
+      "grad_norm": 2.231720080665685,
+      "learning_rate": 9.538956946651816e-06,
+      "loss": 1.1294,
+      "step": 106
+    },
+    {
+      "epoch": 0.6837060702875399,
+      "grad_norm": 1.9603128234576142,
+      "learning_rate": 9.52318163187018e-06,
+      "loss": 1.0876,
+      "step": 107
+    },
+    {
+      "epoch": 0.6900958466453674,
+      "grad_norm": 2.1392812148029736,
+      "learning_rate": 9.507154446436806e-06,
+      "loss": 1.1633,
+      "step": 108
+    },
+    {
+      "epoch": 0.6964856230031949,
+      "grad_norm": 2.272410499944869,
+      "learning_rate": 9.490876282815884e-06,
+      "loss": 1.2247,
+      "step": 109
+    },
+    {
+      "epoch": 0.7028753993610224,
+      "grad_norm": 2.1469318052505706,
+      "learning_rate": 9.474348047447177e-06,
+      "loss": 1.0897,
+      "step": 110
+    },
+    {
+      "epoch": 0.7092651757188498,
+      "grad_norm": 2.159780969055711,
+      "learning_rate": 9.457570660695542e-06,
+      "loss": 1.2255,
+      "step": 111
+    },
+    {
+      "epoch": 0.7156549520766773,
+      "grad_norm": 2.4686968074205726,
+      "learning_rate": 9.440545056799677e-06,
+      "loss": 1.1641,
+      "step": 112
+    },
+    {
+      "epoch": 0.7220447284345048,
+      "grad_norm": 2.2547862843159776,
+      "learning_rate": 9.423272183820109e-06,
+      "loss": 1.1979,
+      "step": 113
+    },
+    {
+      "epoch": 0.7284345047923323,
+      "grad_norm": 2.27724473444956,
+      "learning_rate": 9.405753003586396e-06,
+      "loss": 1.2557,
+      "step": 114
+    },
+    {
+      "epoch": 0.7348242811501597,
+      "grad_norm": 2.5456321939585216,
+      "learning_rate": 9.387988491643558e-06,
+      "loss": 1.1468,
+      "step": 115
+    },
+    {
+      "epoch": 0.7412140575079872,
+      "grad_norm": 2.3676784879006836,
+      "learning_rate": 9.369979637197774e-06,
+      "loss": 1.2817,
+      "step": 116
+    },
+    {
+      "epoch": 0.7476038338658147,
+      "grad_norm": 2.071452107560449,
+      "learning_rate": 9.351727443061284e-06,
+      "loss": 1.2346,
+      "step": 117
+    },
+    {
+      "epoch": 0.7539936102236422,
+      "grad_norm": 1.9877025274939035,
+      "learning_rate": 9.33323292559655e-06,
+      "loss": 1.1517,
+      "step": 118
+    },
+    {
+      "epoch": 0.7603833865814696,
+      "grad_norm": 2.193988573947218,
+      "learning_rate": 9.31449711465967e-06,
+      "loss": 1.1592,
+      "step": 119
+    },
+    {
+      "epoch": 0.7667731629392971,
+      "grad_norm": 1.8767489607648484,
+      "learning_rate": 9.29552105354302e-06,
+      "loss": 1.237,
+      "step": 120
+    },
+    {
+      "epoch": 0.7731629392971247,
+      "grad_norm": 2.1030635199225407,
+      "learning_rate": 9.27630579891716e-06,
+      "loss": 1.2034,
+      "step": 121
+    },
+    {
+      "epoch": 0.7795527156549521,
+      "grad_norm": 2.027092141423695,
+      "learning_rate": 9.256852420771999e-06,
+      "loss": 1.0951,
+      "step": 122
+    },
+    {
+      "epoch": 0.7859424920127795,
+      "grad_norm": 2.1632209630049224,
+      "learning_rate": 9.237162002357214e-06,
+      "loss": 1.1327,
+      "step": 123
+    },
+    {
+      "epoch": 0.792332268370607,
+      "grad_norm": 2.121172839950111,
+      "learning_rate": 9.217235640121927e-06,
+      "loss": 1.2099,
+      "step": 124
+    },
+    {
+      "epoch": 0.7987220447284346,
+      "grad_norm": 2.2253932467304867,
+      "learning_rate": 9.197074443653643e-06,
+      "loss": 1.1512,
+      "step": 125
+    },
+    {
+      "epoch": 0.805111821086262,
+      "grad_norm": 2.0054498620669405,
+      "learning_rate": 9.176679535616477e-06,
+      "loss": 1.1289,
+      "step": 126
+    },
+    {
+      "epoch": 0.8115015974440895,
+      "grad_norm": 2.141568588127354,
+      "learning_rate": 9.156052051688633e-06,
+      "loss": 1.1781,
+      "step": 127
+    },
+    {
+      "epoch": 0.8178913738019169,
+      "grad_norm": 1.9917103164782524,
+      "learning_rate": 9.135193140499155e-06,
+      "loss": 1.303,
+      "step": 128
+    },
+    {
+      "epoch": 0.8242811501597445,
+      "grad_norm": 2.109393580631002,
+      "learning_rate": 9.114103963563986e-06,
+      "loss": 1.218,
+      "step": 129
+    },
+    {
+      "epoch": 0.8306709265175719,
+      "grad_norm": 1.9608467965910812,
+      "learning_rate": 9.092785695221271e-06,
+      "loss": 1.1861,
+      "step": 130
+    },
+    {
+      "epoch": 0.8370607028753994,
+      "grad_norm": 2.2527371323982286,
+      "learning_rate": 9.071239522565978e-06,
+      "loss": 1.1471,
+      "step": 131
+    },
+    {
+      "epoch": 0.8434504792332268,
+      "grad_norm": 2.2882037789195935,
+      "learning_rate": 9.049466645383785e-06,
+      "loss": 1.2044,
+      "step": 132
+    },
+    {
+      "epoch": 0.8498402555910544,
+      "grad_norm": 1.8783108140741698,
+      "learning_rate": 9.027468276084274e-06,
+      "loss": 1.0654,
+      "step": 133
+    },
+    {
+      "epoch": 0.8562300319488818,
+      "grad_norm": 2.181556195900633,
+      "learning_rate": 9.00524563963343e-06,
+      "loss": 1.1876,
+      "step": 134
+    },
+    {
+      "epoch": 0.8626198083067093,
+      "grad_norm": 2.1118568627241214,
+      "learning_rate": 8.982799973485407e-06,
+      "loss": 1.149,
+      "step": 135
+    },
+    {
+      "epoch": 0.8690095846645367,
+      "grad_norm": 1.9977631103473594,
+      "learning_rate": 8.960132527513642e-06,
+      "loss": 1.241,
+      "step": 136
+    },
+    {
+      "epoch": 0.8753993610223643,
+      "grad_norm": 2.175594875693007,
+      "learning_rate": 8.937244563941248e-06,
+      "loss": 1.1798,
+      "step": 137
+    },
+    {
+      "epoch": 0.8817891373801917,
+      "grad_norm": 1.957109658159744,
+      "learning_rate": 8.914137357270723e-06,
+      "loss": 1.1644,
+      "step": 138
+    },
+    {
+      "epoch": 0.8881789137380192,
+      "grad_norm": 2.2821449104582467,
+      "learning_rate": 8.890812194212987e-06,
+      "loss": 1.3045,
+      "step": 139
+    },
+    {
+      "epoch": 0.8945686900958466,
+      "grad_norm": 2.0225222189950713,
+      "learning_rate": 8.867270373615735e-06,
+      "loss": 1.252,
+      "step": 140
+    },
+    {
+      "epoch": 0.9009584664536742,
+      "grad_norm": 2.2060119696696066,
+      "learning_rate": 8.8435132063911e-06,
+      "loss": 1.1079,
+      "step": 141
+    },
+    {
+      "epoch": 0.9073482428115016,
+      "grad_norm": 2.268401775396252,
+      "learning_rate": 8.81954201544267e-06,
+      "loss": 1.1864,
+      "step": 142
+    },
+    {
+      "epoch": 0.9137380191693291,
+      "grad_norm": 2.237757757221249,
+      "learning_rate": 8.79535813559181e-06,
+      "loss": 1.1418,
+      "step": 143
+    },
+    {
+      "epoch": 0.9201277955271565,
+      "grad_norm": 2.365685901904334,
+      "learning_rate": 8.77096291350334e-06,
+      "loss": 1.3222,
+      "step": 144
+    },
+    {
+      "epoch": 0.9265175718849841,
+      "grad_norm": 2.141967663619419,
+      "learning_rate": 8.746357707610544e-06,
+      "loss": 1.1195,
+      "step": 145
+    },
+    {
+      "epoch": 0.9329073482428115,
+      "grad_norm": 2.0672229679280503,
+      "learning_rate": 8.721543888039534e-06,
+      "loss": 1.1288,
+      "step": 146
+    },
+    {
+      "epoch": 0.939297124600639,
+      "grad_norm": 2.0223211301618003,
+      "learning_rate": 8.69652283653294e-06,
+      "loss": 1.2964,
+      "step": 147
+    },
+    {
+      "epoch": 0.9456869009584664,
+      "grad_norm": 2.288439921668011,
+      "learning_rate": 8.671295946372989e-06,
+      "loss": 1.0529,
+      "step": 148
+    },
+    {
+      "epoch": 0.952076677316294,
+      "grad_norm": 2.094215089439829,
+      "learning_rate": 8.6458646223039e-06,
+      "loss": 1.1668,
+      "step": 149
+    },
+    {
+      "epoch": 0.9584664536741214,
+      "grad_norm": 2.036960202702411,
+      "learning_rate": 8.620230280453672e-06,
+      "loss": 1.1601,
+      "step": 150
+    },
+    {
+      "epoch": 0.9648562300319489,
+      "grad_norm": 2.308069594448892,
+      "learning_rate": 8.594394348255239e-06,
+      "loss": 1.2524,
+      "step": 151
+    },
+    {
+      "epoch": 0.9712460063897763,
+      "grad_norm": 2.073184647781397,
+      "learning_rate": 8.568358264366958e-06,
+      "loss": 1.3667,
+      "step": 152
+    },
+    {
+      "epoch": 0.9776357827476039,
+      "grad_norm": 2.6124688635819275,
+      "learning_rate": 8.542123478592518e-06,
+      "loss": 1.0973,
+      "step": 153
+    },
+    {
+      "epoch": 0.9840255591054313,
+      "grad_norm": 2.023174608703274,
+      "learning_rate": 8.515691451800206e-06,
+      "loss": 1.2798,
+      "step": 154
+    },
+    {
+      "epoch": 0.9904153354632588,
+      "grad_norm": 2.1839751935320963,
+      "learning_rate": 8.489063655841552e-06,
+      "loss": 1.1375,
+      "step": 155
+    },
+    {
+      "epoch": 0.9968051118210862,
+      "grad_norm": 2.114621211140617,
+      "learning_rate": 8.462241573469378e-06,
+      "loss": 0.9936,
+      "step": 156
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.114621211140617,
+      "learning_rate": 8.435226698255228e-06,
+      "loss": 1.137,
+      "step": 157
+    },
+    {
+      "epoch": 1.0063897763578276,
+      "grad_norm": 3.4867497812189288,
+      "learning_rate": 8.408020534506195e-06,
+      "loss": 0.8376,
+      "step": 158
+    },
+    {
+      "epoch": 1.012779552715655,
+      "grad_norm": 2.5440510855222676,
+      "learning_rate": 8.380624597181165e-06,
+      "loss": 0.8312,
+      "step": 159
+    },
+    {
+      "epoch": 1.0191693290734825,
+      "grad_norm": 2.2596612524017186,
+      "learning_rate": 8.353040411806449e-06,
+      "loss": 0.7729,
+      "step": 160
+    },
+    {
+      "epoch": 1.0255591054313098,
+      "grad_norm": 1.8463551184455353,
+      "learning_rate": 8.325269514390835e-06,
+      "loss": 0.8626,
+      "step": 161
+    },
+    {
+      "epoch": 1.0319488817891374,
+      "grad_norm": 2.0882217693894813,
+      "learning_rate": 8.297313451340064e-06,
+      "loss": 0.7676,
+      "step": 162
+    },
+    {
+      "epoch": 1.038338658146965,
+      "grad_norm": 2.0883744868398226,
+      "learning_rate": 8.269173779370712e-06,
+      "loss": 0.9419,
+      "step": 163
+    },
+    {
+      "epoch": 1.0447284345047922,
+      "grad_norm": 1.7943635404247467,
+      "learning_rate": 8.240852065423507e-06,
+      "loss": 0.7972,
+      "step": 164
+    },
+    {
+      "epoch": 1.0511182108626198,
+      "grad_norm": 1.9220993590643272,
+      "learning_rate": 8.21234988657607e-06,
+      "loss": 0.8412,
+      "step": 165
+    },
+    {
+      "epoch": 1.0575079872204474,
+      "grad_norm": 1.8676951285327783,
+      "learning_rate": 8.183668829955111e-06,
+      "loss": 0.8471,
+      "step": 166
+    },
+    {
+      "epoch": 1.0638977635782747,
+      "grad_norm": 1.9115531535231205,
+      "learning_rate": 8.154810492648038e-06,
+      "loss": 0.8021,
+      "step": 167
+    },
+    {
+      "epoch": 1.0702875399361023,
+      "grad_norm": 1.732608310176797,
+      "learning_rate": 8.125776481614025e-06,
+      "loss": 0.6606,
+      "step": 168
+    },
+    {
+      "epoch": 1.0766773162939298,
+      "grad_norm": 1.9131580562412362,
+      "learning_rate": 8.096568413594533e-06,
+      "loss": 0.7635,
+      "step": 169
+    },
+    {
+      "epoch": 1.0830670926517572,
+      "grad_norm": 1.8959582550313192,
+      "learning_rate": 8.067187915023283e-06,
+      "loss": 0.6924,
+      "step": 170
+    },
+    {
+      "epoch": 1.0894568690095847,
+      "grad_norm": 2.162226106303363,
+      "learning_rate": 8.037636621935686e-06,
+      "loss": 0.7722,
+      "step": 171
+    },
+    {
+      "epoch": 1.095846645367412,
+      "grad_norm": 2.387269781570661,
+      "learning_rate": 8.007916179877742e-06,
+      "loss": 0.7872,
+      "step": 172
+    },
+    {
+      "epoch": 1.1022364217252396,
+      "grad_norm": 1.8369220466875913,
+      "learning_rate": 7.978028243814416e-06,
+      "loss": 0.7746,
+      "step": 173
+    },
+    {
+      "epoch": 1.1086261980830672,
+      "grad_norm": 2.3362355498895853,
+      "learning_rate": 7.947974478037468e-06,
+      "loss": 0.7845,
+      "step": 174
+    },
+    {
+      "epoch": 1.1150159744408945,
+      "grad_norm": 2.0503755528528553,
+      "learning_rate": 7.917756556072792e-06,
+      "loss": 0.7205,
+      "step": 175
+    },
+    {
+      "epoch": 1.121405750798722,
+      "grad_norm": 2.165347300861018,
+      "learning_rate": 7.887376160587214e-06,
+      "loss": 0.8152,
+      "step": 176
+    },
+    {
+      "epoch": 1.1277955271565494,
+      "grad_norm": 2.2635558547519747,
+      "learning_rate": 7.85683498329481e-06,
+      "loss": 0.6753,
+      "step": 177
+    },
+    {
+      "epoch": 1.134185303514377,
+      "grad_norm": 1.9110658352060157,
+      "learning_rate": 7.826134724862687e-06,
+      "loss": 0.825,
+      "step": 178
+    },
+    {
+      "epoch": 1.1405750798722045,
+      "grad_norm": 2.2263812309151825,
+      "learning_rate": 7.795277094816292e-06,
+      "loss": 0.8166,
+      "step": 179
+    },
+    {
+      "epoch": 1.1469648562300319,
+      "grad_norm": 1.8575499012395842,
+      "learning_rate": 7.764263811444214e-06,
+      "loss": 0.831,
+      "step": 180
+    },
+    {
+      "epoch": 1.1533546325878594,
+      "grad_norm": 2.360397566704423,
+      "learning_rate": 7.733096601702508e-06,
+      "loss": 0.8137,
+      "step": 181
+    },
+    {
+      "epoch": 1.159744408945687,
+      "grad_norm": 2.2444203392924895,
+      "learning_rate": 7.70177720111852e-06,
+      "loss": 0.6924,
+      "step": 182
+    },
+    {
+      "epoch": 1.1661341853035143,
+      "grad_norm": 1.6246406063388128,
+      "learning_rate": 7.67030735369426e-06,
+      "loss": 0.7463,
+      "step": 183
+    },
+    {
+      "epoch": 1.1725239616613419,
+      "grad_norm": 2.173100756669063,
+      "learning_rate": 7.638688811809274e-06,
+      "loss": 0.7659,
+      "step": 184
+    },
+    {
+      "epoch": 1.1789137380191694,
+      "grad_norm": 2.0758535951181325,
+      "learning_rate": 7.6069233361230696e-06,
+      "loss": 0.8334,
+      "step": 185
+    },
+    {
+      "epoch": 1.1853035143769968,
+      "grad_norm": 2.0584161403900088,
+      "learning_rate": 7.575012695477076e-06,
+      "loss": 0.7935,
+      "step": 186
+    },
+    {
+      "epoch": 1.1916932907348243,
+      "grad_norm": 2.1564370807060524,
+      "learning_rate": 7.542958666796149e-06,
+      "loss": 0.7672,
+      "step": 187
+    },
+    {
+      "epoch": 1.1980830670926517,
+      "grad_norm": 1.8307453253127797,
+      "learning_rate": 7.510763034989616e-06,
+      "loss": 0.6783,
+      "step": 188
+    },
+    {
+      "epoch": 1.2044728434504792,
+      "grad_norm": 1.8290500329133041,
+      "learning_rate": 7.478427592851894e-06,
+      "loss": 0.6824,
+      "step": 189
+    },
+    {
+      "epoch": 1.2108626198083068,
+      "grad_norm": 2.0174580703448037,
+      "learning_rate": 7.44595414096265e-06,
+      "loss": 0.7652,
+      "step": 190
+    },
+    {
+      "epoch": 1.2172523961661341,
+      "grad_norm": 1.9899638513663098,
+      "learning_rate": 7.413344487586542e-06,
+      "loss": 0.7497,
+      "step": 191
+    },
+    {
+      "epoch": 1.2236421725239617,
+      "grad_norm": 2.4723485730809243,
+      "learning_rate": 7.380600448572532e-06,
+      "loss": 0.8049,
+      "step": 192
+    },
+    {
+      "epoch": 1.230031948881789,
+      "grad_norm": 2.034861515675997,
+      "learning_rate": 7.347723847252756e-06,
+      "loss": 0.8709,
+      "step": 193
+    },
+    {
+      "epoch": 1.2364217252396166,
+      "grad_norm": 2.138975439198192,
+      "learning_rate": 7.314716514341007e-06,
+      "loss": 0.6953,
+      "step": 194
+    },
+    {
+      "epoch": 1.2428115015974441,
+      "grad_norm": 1.9632002426077662,
+      "learning_rate": 7.28158028783079e-06,
+      "loss": 0.7643,
+      "step": 195
+    },
+    {
+      "epoch": 1.2492012779552715,
+      "grad_norm": 2.1518741495487625,
+      "learning_rate": 7.248317012892969e-06,
+      "loss": 0.7793,
+      "step": 196
+    },
+    {
+      "epoch": 1.255591054313099,
+      "grad_norm": 2.449964832983734,
+      "learning_rate": 7.214928541773027e-06,
+      "loss": 0.7526,
+      "step": 197
+    },
+    {
+      "epoch": 1.2619808306709266,
+      "grad_norm": 2.069692679407715,
+      "learning_rate": 7.1814167336879195e-06,
+      "loss": 0.6993,
+      "step": 198
+    },
+    {
+      "epoch": 1.268370607028754,
+      "grad_norm": 2.1802916449943734,
+      "learning_rate": 7.147783454722545e-06,
+      "loss": 0.5996,
+      "step": 199
+    },
+    {
+      "epoch": 1.2747603833865815,
+      "grad_norm": 2.1168439959277086,
+      "learning_rate": 7.1140305777258355e-06,
+      "loss": 0.8294,
+      "step": 200
+    },
+    {
+      "epoch": 1.281150159744409,
+      "grad_norm": 1.8184694551749518,
+      "learning_rate": 7.080159982206471e-06,
+      "loss": 0.6787,
+      "step": 201
+    },
+    {
+      "epoch": 1.2875399361022364,
+      "grad_norm": 1.902418537487257,
+      "learning_rate": 7.046173554228213e-06,
+      "loss": 0.8076,
+      "step": 202
+    },
+    {
+      "epoch": 1.293929712460064,
+      "grad_norm": 1.9892137836089137,
+      "learning_rate": 7.012073186304885e-06,
+      "loss": 0.8541,
+      "step": 203
+    },
+    {
+      "epoch": 1.3003194888178915,
+      "grad_norm": 2.3335747045667903,
+      "learning_rate": 6.9778607772949894e-06,
+      "loss": 0.7712,
+      "step": 204
+    },
+    {
+      "epoch": 1.3067092651757188,
+      "grad_norm": 1.983789823335996,
+      "learning_rate": 6.943538232295965e-06,
+      "loss": 0.7885,
+      "step": 205
+    },
+    {
+      "epoch": 1.3130990415335464,
+      "grad_norm": 2.149089239306387,
+      "learning_rate": 6.909107462538113e-06,
+      "loss": 0.786,
+      "step": 206
+    },
+    {
+      "epoch": 1.3194888178913737,
+      "grad_norm": 2.3740320043022356,
+      "learning_rate": 6.874570385278161e-06,
+      "loss": 0.6849,
+      "step": 207
+    },
+    {
+      "epoch": 1.3258785942492013,
+      "grad_norm": 2.0274327595584727,
+      "learning_rate": 6.839928923692505e-06,
+      "loss": 0.6331,
+      "step": 208
+    },
+    {
+      "epoch": 1.3322683706070286,
+      "grad_norm": 1.9834070538746025,
+      "learning_rate": 6.805185006770125e-06,
+      "loss": 0.7026,
+      "step": 209
+    },
+    {
+      "epoch": 1.3386581469648562,
+      "grad_norm": 1.8332634885192054,
+      "learning_rate": 6.7703405692051585e-06,
+      "loss": 0.7521,
+      "step": 210
+    },
+    {
+      "epoch": 1.3450479233226837,
+      "grad_norm": 2.3224931980911303,
+      "learning_rate": 6.735397551289179e-06,
+      "loss": 0.8352,
+      "step": 211
+    },
+    {
+      "epoch": 1.351437699680511,
+      "grad_norm": 2.5988749953217383,
+      "learning_rate": 6.700357898803146e-06,
+      "loss": 0.7261,
+      "step": 212
+    },
+    {
+      "epoch": 1.3578274760383386,
+      "grad_norm": 2.0025941890434518,
+      "learning_rate": 6.665223562909058e-06,
+      "loss": 0.752,
+      "step": 213
+    },
+    {
+      "epoch": 1.3642172523961662,
+      "grad_norm": 1.8512272814617676,
+      "learning_rate": 6.629996500041299e-06,
+      "loss": 0.7175,
+      "step": 214
+    },
+    {
+      "epoch": 1.3706070287539935,
+      "grad_norm": 2.221397828681688,
+      "learning_rate": 6.5946786717977026e-06,
+      "loss": 0.7518,
+      "step": 215
+    },
+    {
+      "epoch": 1.376996805111821,
+      "grad_norm": 2.0249123351492155,
+      "learning_rate": 6.5592720448303174e-06,
+      "loss": 0.8235,
+      "step": 216
+    },
+    {
+      "epoch": 1.3833865814696487,
+      "grad_norm": 2.18979667251631,
+      "learning_rate": 6.523778590735892e-06,
+      "loss": 0.9315,
+      "step": 217
+    },
+    {
+      "epoch": 1.389776357827476,
+      "grad_norm": 1.8960015883246881,
+      "learning_rate": 6.488200285946094e-06,
+      "loss": 0.7076,
+      "step": 218
+    },
+    {
+      "epoch": 1.3961661341853036,
+      "grad_norm": 1.8352798094679064,
+      "learning_rate": 6.452539111617454e-06,
+      "loss": 0.7501,
+      "step": 219
+    },
+    {
+      "epoch": 1.4025559105431311,
+      "grad_norm": 2.097097886969805,
+      "learning_rate": 6.416797053521039e-06,
+      "loss": 0.7881,
+      "step": 220
+    },
+    {
+      "epoch": 1.4089456869009584,
+      "grad_norm": 1.9059659683647745,
+      "learning_rate": 6.380976101931879e-06,
+      "loss": 0.7922,
+      "step": 221
+    },
+    {
+      "epoch": 1.415335463258786,
+      "grad_norm": 2.1954958084121965,
+      "learning_rate": 6.345078251518144e-06,
+      "loss": 0.8997,
+      "step": 222
+    },
+    {
+      "epoch": 1.4217252396166133,
+      "grad_norm": 1.966127388072899,
+      "learning_rate": 6.3091055012300675e-06,
+      "loss": 0.7675,
+      "step": 223
+    },
+    {
+      "epoch": 1.428115015974441,
+      "grad_norm": 2.3361083036828094,
+      "learning_rate": 6.273059854188636e-06,
+      "loss": 0.8397,
+      "step": 224
+    },
+    {
+      "epoch": 1.4345047923322682,
+      "grad_norm": 2.139956577956786,
+      "learning_rate": 6.236943317574054e-06,
+      "loss": 0.7714,
+      "step": 225
+    },
+    {
+      "epoch": 1.4408945686900958,
+      "grad_norm": 1.9866600216210257,
+      "learning_rate": 6.200757902513962e-06,
+      "loss": 0.8059,
+      "step": 226
+    },
+    {
+      "epoch": 1.4472843450479234,
+      "grad_norm": 2.259805416277339,
+      "learning_rate": 6.164505623971458e-06,
+      "loss": 0.802,
+      "step": 227
+    },
+    {
+      "epoch": 1.4536741214057507,
+      "grad_norm": 1.8782576150494317,
+      "learning_rate": 6.128188500632892e-06,
+      "loss": 0.7065,
+      "step": 228
+    },
+    {
+      "epoch": 1.4600638977635783,
+      "grad_norm": 2.1548044609256314,
+      "learning_rate": 6.091808554795462e-06,
+      "loss": 0.853,
+      "step": 229
+    },
+    {
+      "epoch": 1.4664536741214058,
+      "grad_norm": 2.1364715239439978,
+      "learning_rate": 6.055367812254592e-06,
+      "loss": 0.8676,
+      "step": 230
+    },
+    {
+      "epoch": 1.4728434504792332,
+      "grad_norm": 1.9045652390431531,
+      "learning_rate": 6.0188683021911394e-06,
+      "loss": 0.8338,
+      "step": 231
+    },
+    {
+      "epoch": 1.4792332268370607,
+      "grad_norm": 2.10885335410803,
+      "learning_rate": 5.982312057058392e-06,
+      "loss": 0.829,
+      "step": 232
+    },
+    {
+      "epoch": 1.4856230031948883,
+      "grad_norm": 2.1881709571010486,
+      "learning_rate": 5.9457011124689025e-06,
+      "loss": 0.9104,
+      "step": 233
+    },
+    {
+      "epoch": 1.4920127795527156,
+      "grad_norm": 2.3768700262231093,
+      "learning_rate": 5.9090375070811215e-06,
+      "loss": 0.6876,
+      "step": 234
+    },
+    {
+      "epoch": 1.4984025559105432,
+      "grad_norm": 2.261550292871675,
+      "learning_rate": 5.872323282485889e-06,
+      "loss": 0.7453,
+      "step": 235
+    },
+    {
+      "epoch": 1.5047923322683707,
+      "grad_norm": 1.9988487840521534,
+      "learning_rate": 5.835560483092743e-06,
+      "loss": 0.8449,
+      "step": 236
+    },
+    {
+      "epoch": 1.511182108626198,
+      "grad_norm": 2.0224359380963652,
+      "learning_rate": 5.798751156016085e-06,
+      "loss": 0.8223,
+      "step": 237
+    },
+    {
+      "epoch": 1.5175718849840254,
+      "grad_norm": 2.062039968011777,
+      "learning_rate": 5.7618973509611755e-06,
+      "loss": 0.8195,
+      "step": 238
+    },
+    {
+      "epoch": 1.5239616613418532,
+      "grad_norm": 1.6157234928139612,
+      "learning_rate": 5.72500112011001e-06,
+      "loss": 0.7592,
+      "step": 239
+    },
+    {
+      "epoch": 1.5303514376996805,
+      "grad_norm": 1.9080273993370855,
+      "learning_rate": 5.688064518007036e-06,
+      "loss": 0.7381,
+      "step": 240
+    },
+    {
+      "epoch": 1.5367412140575079,
+      "grad_norm": 2.2625716403086344,
+      "learning_rate": 5.651089601444752e-06,
+      "loss": 0.8195,
+      "step": 241
+    },
+    {
+      "epoch": 1.5431309904153354,
+      "grad_norm": 2.0895666610527126,
+      "learning_rate": 5.614078429349172e-06,
+      "loss": 0.7222,
+      "step": 242
+    },
+    {
+      "epoch": 1.549520766773163,
+      "grad_norm": 1.99029735584039,
+      "learning_rate": 5.577033062665179e-06,
+      "loss": 0.838,
+      "step": 243
+    },
+    {
+      "epoch": 1.5559105431309903,
+      "grad_norm": 2.1194831420040985,
+      "learning_rate": 5.53995556424176e-06,
+      "loss": 0.707,
+      "step": 244
+    },
+    {
+      "epoch": 1.5623003194888179,
+      "grad_norm": 2.3137301949010984,
+      "learning_rate": 5.50284799871714e-06,
+      "loss": 0.8003,
+      "step": 245
+    },
+    {
+      "epoch": 1.5686900958466454,
+      "grad_norm": 2.2464503074012674,
+      "learning_rate": 5.465712432403812e-06,
+      "loss": 0.8576,
+      "step": 246
+    },
+    {
+      "epoch": 1.5750798722044728,
+      "grad_norm": 2.320906889466704,
+      "learning_rate": 5.428550933173476e-06,
+      "loss": 0.916,
+      "step": 247
+    },
+    {
+      "epoch": 1.5814696485623003,
+      "grad_norm": 2.099438841370686,
+      "learning_rate": 5.391365570341893e-06,
+      "loss": 0.7929,
+      "step": 248
+    },
+    {
+      "epoch": 1.5878594249201279,
+      "grad_norm": 1.859084494736742,
+      "learning_rate": 5.3541584145536475e-06,
+      "loss": 0.6533,
+      "step": 249
+    },
+    {
+      "epoch": 1.5942492012779552,
+      "grad_norm": 2.278505423549853,
+      "learning_rate": 5.3169315376668566e-06,
+      "loss": 0.7552,
+      "step": 250
+    },
+    {
+      "epoch": 1.6006389776357828,
+      "grad_norm": 1.6860670526764372,
+      "learning_rate": 5.279687012637798e-06,
+      "loss": 0.8399,
+      "step": 251
+    },
+    {
+      "epoch": 1.6070287539936103,
+      "grad_norm": 2.335897193410633,
+      "learning_rate": 5.242426913405471e-06,
+      "loss": 0.7687,
+      "step": 252
+    },
+    {
+      "epoch": 1.6134185303514377,
+      "grad_norm": 2.060267209744552,
+      "learning_rate": 5.2051533147761155e-06,
+      "loss": 0.7508,
+      "step": 253
+    },
+    {
+      "epoch": 1.619808306709265,
+      "grad_norm": 2.1497401104875373,
+      "learning_rate": 5.167868292307679e-06,
+      "loss": 0.7592,
+      "step": 254
+    },
+    {
+      "epoch": 1.6261980830670928,
+      "grad_norm": 1.7903715915769907,
+      "learning_rate": 5.130573922194236e-06,
+      "loss": 0.801,
+      "step": 255
+    },
+    {
+      "epoch": 1.6325878594249201,
+      "grad_norm": 1.994091783861481,
+      "learning_rate": 5.093272281150383e-06,
+      "loss": 0.7861,
+      "step": 256
+    },
+    {
+      "epoch": 1.6389776357827475,
+      "grad_norm": 2.131683372343815,
+      "learning_rate": 5.05596544629559e-06,
+      "loss": 0.8001,
+      "step": 257
+    },
+    {
+      "epoch": 1.645367412140575,
+      "grad_norm": 2.1302879758896855,
+      "learning_rate": 5.018655495038542e-06,
+      "loss": 0.8043,
+      "step": 258
+    },
+    {
+      "epoch": 1.6517571884984026,
+      "grad_norm": 1.8253111170736016,
+      "learning_rate": 4.981344504961459e-06,
+      "loss": 0.8891,
+      "step": 259
+    },
+    {
+      "epoch": 1.65814696485623,
+      "grad_norm": 1.9749425046329303,
+      "learning_rate": 4.944034553704412e-06,
+      "loss": 0.8087,
+      "step": 260
+    },
+    {
+      "epoch": 1.6645367412140575,
+      "grad_norm": 2.193811659404884,
+      "learning_rate": 4.906727718849619e-06,
+      "loss": 0.7552,
+      "step": 261
+    },
+    {
+      "epoch": 1.670926517571885,
+      "grad_norm": 2.1765708924053575,
+      "learning_rate": 4.8694260778057655e-06,
+      "loss": 0.7897,
+      "step": 262
+    },
+    {
+      "epoch": 1.6773162939297124,
+      "grad_norm": 2.0608183940458398,
+      "learning_rate": 4.832131707692322e-06,
+      "loss": 0.7122,
+      "step": 263
+    },
+    {
+      "epoch": 1.68370607028754,
+      "grad_norm": 2.1978305930011555,
+      "learning_rate": 4.7948466852238844e-06,
+      "loss": 0.8197,
+      "step": 264
+    },
+    {
+      "epoch": 1.6900958466453675,
+      "grad_norm": 2.299856637633407,
+      "learning_rate": 4.757573086594529e-06,
+      "loss": 0.9011,
+      "step": 265
+    },
+    {
+      "epoch": 1.6964856230031948,
+      "grad_norm": 2.1505800913123707,
+      "learning_rate": 4.720312987362204e-06,
+      "loss": 0.784,
+      "step": 266
+    },
+    {
+      "epoch": 1.7028753993610224,
+      "grad_norm": 2.1562863338865714,
+      "learning_rate": 4.683068462333144e-06,
+      "loss": 0.8057,
+      "step": 267
+    },
+    {
+      "epoch": 1.70926517571885,
+      "grad_norm": 2.284446615571358,
+      "learning_rate": 4.645841585446356e-06,
+      "loss": 0.8029,
+      "step": 268
+    },
+    {
+      "epoch": 1.7156549520766773,
+      "grad_norm": 2.0510265573141413,
+      "learning_rate": 4.6086344296581095e-06,
+      "loss": 0.7218,
+      "step": 269
+    },
+    {
+      "epoch": 1.7220447284345048,
+      "grad_norm": 2.1687340523353353,
+      "learning_rate": 4.5714490668265245e-06,
+      "loss": 0.9969,
+      "step": 270
+    },
+    {
+      "epoch": 1.7284345047923324,
+      "grad_norm": 2.0736960494924923,
+      "learning_rate": 4.534287567596189e-06,
+      "loss": 0.85,
+      "step": 271
+    },
+    {
+      "epoch": 1.7348242811501597,
+      "grad_norm": 1.9770333120267458,
+      "learning_rate": 4.497152001282861e-06,
+      "loss": 0.8745,
+      "step": 272
+    },
+    {
+      "epoch": 1.741214057507987,
+      "grad_norm": 2.084503899334248,
+      "learning_rate": 4.460044435758241e-06,
+      "loss": 0.788,
+      "step": 273
+    },
+    {
+      "epoch": 1.7476038338658149,
+      "grad_norm": 2.0686608595640084,
+      "learning_rate": 4.4229669373348225e-06,
+      "loss": 0.7217,
+      "step": 274
+    },
+    {
+      "epoch": 1.7539936102236422,
+      "grad_norm": 1.8726775800589226,
+      "learning_rate": 4.3859215706508295e-06,
+      "loss": 0.9332,
+      "step": 275
+    },
+    {
+      "epoch": 1.7603833865814695,
+      "grad_norm": 2.321014760523366,
+      "learning_rate": 4.348910398555249e-06,
+      "loss": 0.848,
+      "step": 276
+    },
+    {
+      "epoch": 1.766773162939297,
+      "grad_norm": 2.1582714047829903,
+      "learning_rate": 4.311935481992965e-06,
+      "loss": 0.6687,
+      "step": 277
+    },
+    {
+      "epoch": 1.7731629392971247,
+      "grad_norm": 2.0830550177571796,
+      "learning_rate": 4.274998879889991e-06,
+      "loss": 0.7575,
+      "step": 278
+    },
+    {
+      "epoch": 1.779552715654952,
+      "grad_norm": 2.1028578941021845,
+      "learning_rate": 4.238102649038825e-06,
+      "loss": 0.7409,
+      "step": 279
+    },
+    {
+      "epoch": 1.7859424920127795,
+      "grad_norm": 1.8878312811111115,
+      "learning_rate": 4.2012488439839185e-06,
+      "loss": 0.8274,
+      "step": 280
+    },
+    {
+      "epoch": 1.792332268370607,
+      "grad_norm": 2.291366102734256,
+      "learning_rate": 4.164439516907258e-06,
+      "loss": 0.7611,
+      "step": 281
+    },
+    {
+      "epoch": 1.7987220447284344,
+      "grad_norm": 1.916507380045079,
+      "learning_rate": 4.127676717514114e-06,
+      "loss": 0.5988,
+      "step": 282
+    },
+    {
+      "epoch": 1.805111821086262,
+      "grad_norm": 2.2720293658938027,
+      "learning_rate": 4.090962492918881e-06,
+      "loss": 0.8268,
+      "step": 283
+    },
+    {
+      "epoch": 1.8115015974440896,
+      "grad_norm": 1.9518349825587402,
+      "learning_rate": 4.054298887531099e-06,
+      "loss": 0.7334,
+      "step": 284
+    },
+    {
+      "epoch": 1.817891373801917,
+      "grad_norm": 2.140363550906506,
+      "learning_rate": 4.017687942941609e-06,
+      "loss": 0.7335,
+      "step": 285
+    },
+    {
+      "epoch": 1.8242811501597445,
+      "grad_norm": 1.9883426651308305,
+      "learning_rate": 3.981131697808862e-06,
+      "loss": 0.828,
+      "step": 286
+    },
+    {
+      "epoch": 1.830670926517572,
+      "grad_norm": 1.951493400947361,
+      "learning_rate": 3.94463218774541e-06,
+      "loss": 0.6701,
+      "step": 287
+    },
+    {
+      "epoch": 1.8370607028753994,
+      "grad_norm": 1.714247520762039,
+      "learning_rate": 3.90819144520454e-06,
+      "loss": 0.7093,
+      "step": 288
+    },
+    {
+      "epoch": 1.8434504792332267,
+      "grad_norm": 2.0546468817069505,
+      "learning_rate": 3.8718114993671086e-06,
+      "loss": 0.777,
+      "step": 289
+    },
+    {
+      "epoch": 1.8498402555910545,
+      "grad_norm": 2.0974753085677222,
+      "learning_rate": 3.835494376028544e-06,
+      "loss": 0.7093,
+      "step": 290
+    },
+    {
+      "epoch": 1.8562300319488818,
+      "grad_norm": 2.286892348662737,
+      "learning_rate": 3.799242097486038e-06,
+      "loss": 0.8641,
+      "step": 291
+    },
+    {
+      "epoch": 1.8626198083067091,
+      "grad_norm": 2.1470027841795614,
+      "learning_rate": 3.7630566824259456e-06,
+      "loss": 0.7596,
+      "step": 292
+    },
+    {
+      "epoch": 1.8690095846645367,
+      "grad_norm": 2.1876379291519887,
+      "learning_rate": 3.726940145811363e-06,
+      "loss": 0.9124,
+      "step": 293
+    },
+    {
+      "epoch": 1.8753993610223643,
+      "grad_norm": 2.073070287437658,
+      "learning_rate": 3.6908944987699346e-06,
+      "loss": 0.7598,
+      "step": 294
+    },
+    {
+      "epoch": 1.8817891373801916,
+      "grad_norm": 2.1569863741791466,
+      "learning_rate": 3.6549217484818576e-06,
+      "loss": 0.7232,
+      "step": 295
+    },
+    {
+      "epoch": 1.8881789137380192,
+      "grad_norm": 2.041941562515545,
+      "learning_rate": 3.6190238980681235e-06,
+      "loss": 0.6949,
+      "step": 296
+    },
+    {
+      "epoch": 1.8945686900958467,
+      "grad_norm": 2.30271960588349,
+      "learning_rate": 3.583202946478963e-06,
+      "loss": 0.8439,
+      "step": 297
+    },
+    {
+      "epoch": 1.900958466453674,
+      "grad_norm": 2.0264437896681207,
+      "learning_rate": 3.5474608883825475e-06,
+      "loss": 0.7475,
+      "step": 298
+    },
+    {
+      "epoch": 1.9073482428115016,
+      "grad_norm": 1.9457581984726628,
+      "learning_rate": 3.5117997140539073e-06,
+      "loss": 0.7066,
+      "step": 299
+    },
+    {
+      "epoch": 1.9137380191693292,
+      "grad_norm": 1.9462959642151252,
+      "learning_rate": 3.47622140926411e-06,
+      "loss": 0.7308,
+      "step": 300
+    },
+    {
+      "epoch": 1.9201277955271565,
+      "grad_norm": 2.092610478552592,
+      "learning_rate": 3.4407279551696846e-06,
+      "loss": 0.7872,
+      "step": 301
+    },
+    {
+      "epoch": 1.926517571884984,
+      "grad_norm": 1.8472874883400652,
+      "learning_rate": 3.4053213282022983e-06,
+      "loss": 0.7524,
+      "step": 302
+    },
+    {
+      "epoch": 1.9329073482428116,
+      "grad_norm": 1.901162640597834,
+      "learning_rate": 3.370003499958703e-06,
+      "loss": 0.7164,
+      "step": 303
+    },
+    {
+      "epoch": 1.939297124600639,
+      "grad_norm": 1.8688778106140063,
+      "learning_rate": 3.334776437090944e-06,
+      "loss": 0.7011,
+      "step": 304
+    },
+    {
+      "epoch": 1.9456869009584663,
+      "grad_norm": 2.356897701654345,
+      "learning_rate": 3.2996421011968546e-06,
+      "loss": 0.6435,
+      "step": 305
+    },
+    {
+      "epoch": 1.952076677316294,
+      "grad_norm": 2.523583795757241,
+      "learning_rate": 3.264602448710822e-06,
+      "loss": 0.76,
+      "step": 306
+    },
+    {
+      "epoch": 1.9584664536741214,
+      "grad_norm": 1.9574404195601653,
+      "learning_rate": 3.2296594307948428e-06,
+      "loss": 0.8161,
+      "step": 307
+    },
+    {
+      "epoch": 1.9648562300319488,
+      "grad_norm": 2.15492898863906,
+      "learning_rate": 3.194814993229878e-06,
+      "loss": 0.7746,
+      "step": 308
+    },
+    {
+      "epoch": 1.9712460063897763,
+      "grad_norm": 2.010578131919025,
+      "learning_rate": 3.1600710763074972e-06,
+      "loss": 0.8067,
+      "step": 309
+    },
+    {
+      "epoch": 1.9776357827476039,
+      "grad_norm": 1.991937348455583,
+      "learning_rate": 3.125429614721842e-06,
+      "loss": 0.6662,
+      "step": 310
+    },
+    {
+      "epoch": 1.9840255591054312,
+      "grad_norm": 2.124227953910074,
+      "learning_rate": 3.090892537461889e-06,
+      "loss": 0.8373,
+      "step": 311
+    },
+    {
+      "epoch": 1.9904153354632588,
+      "grad_norm": 1.9267525903621767,
+      "learning_rate": 3.056461767704037e-06,
+      "loss": 0.7735,
+      "step": 312
+    },
+    {
+      "epoch": 1.9968051118210863,
+      "grad_norm": 2.176290715664279,
+      "learning_rate": 3.0221392227050126e-06,
+      "loss": 0.83,
+      "step": 313
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.1455065522489942,
+      "learning_rate": 2.9879268136951163e-06,
+      "loss": 0.6453,
+      "step": 314
+    },
+    {
+      "epoch": 2.0063897763578273,
+      "grad_norm": 3.1822765413878225,
+      "learning_rate": 2.953826445771788e-06,
+      "loss": 0.437,
+      "step": 315
+    },
+    {
+      "epoch": 2.012779552715655,
+      "grad_norm": 3.366770013317489,
+      "learning_rate": 2.9198400177935303e-06,
+      "loss": 0.4245,
+      "step": 316
+    },
+    {
+      "epoch": 2.0191693290734825,
+      "grad_norm": 3.26466588175627,
+      "learning_rate": 2.8859694222741653e-06,
+      "loss": 0.4096,
+      "step": 317
+    },
+    {
+      "epoch": 2.02555910543131,
+      "grad_norm": 3.1916614763418494,
+      "learning_rate": 2.852216545277456e-06,
+      "loss": 0.5049,
+      "step": 318
+    },
+    {
+      "epoch": 2.0319488817891376,
+      "grad_norm": 2.4873488239851227,
+      "learning_rate": 2.8185832663120817e-06,
+      "loss": 0.491,
+      "step": 319
+    },
+    {
+      "epoch": 2.038338658146965,
+      "grad_norm": 2.4295793465904474,
+      "learning_rate": 2.785071458226972e-06,
+      "loss": 0.4575,
+      "step": 320
+    },
+    {
+      "epoch": 2.0447284345047922,
+      "grad_norm": 2.264908078075383,
+      "learning_rate": 2.7516829871070295e-06,
+      "loss": 0.5299,
+      "step": 321
+    },
+    {
+      "epoch": 2.0511182108626196,
+      "grad_norm": 2.115450964357911,
+      "learning_rate": 2.718419712169213e-06,
+      "loss": 0.5705,
+      "step": 322
+    },
+    {
+      "epoch": 2.0575079872204474,
+      "grad_norm": 2.1328168895257384,
+      "learning_rate": 2.685283485658995e-06,
+      "loss": 0.4552,
+      "step": 323
+    },
+    {
+      "epoch": 2.0638977635782747,
+      "grad_norm": 2.522885900225875,
+      "learning_rate": 2.6522761527472464e-06,
+      "loss": 0.423,
+      "step": 324
+    },
+    {
+      "epoch": 2.070287539936102,
+      "grad_norm": 3.677486298025955,
+      "learning_rate": 2.6193995514274705e-06,
+      "loss": 0.5114,
+      "step": 325
+    },
+    {
+      "epoch": 2.07667731629393,
+      "grad_norm": 3.147438831845241,
+      "learning_rate": 2.586655512413458e-06,
+      "loss": 0.5074,
+      "step": 326
+    },
+    {
+      "epoch": 2.083067092651757,
+      "grad_norm": 3.454909400135248,
+      "learning_rate": 2.554045859037353e-06,
+      "loss": 0.5319,
+      "step": 327
+    },
+    {
+      "epoch": 2.0894568690095845,
+      "grad_norm": 3.2027788217049293,
+      "learning_rate": 2.521572407148107e-06,
+      "loss": 0.4018,
+      "step": 328
+    },
+    {
+      "epoch": 2.0958466453674123,
+      "grad_norm": 2.82713510067431,
+      "learning_rate": 2.4892369650103837e-06,
+      "loss": 0.4096,
+      "step": 329
+    },
+    {
+      "epoch": 2.1022364217252396,
+      "grad_norm": 3.3775502635423846,
+      "learning_rate": 2.4570413332038523e-06,
+      "loss": 0.4084,
+      "step": 330
+    },
+    {
+      "epoch": 2.108626198083067,
+      "grad_norm": 2.5912870135939854,
+      "learning_rate": 2.4249873045229244e-06,
+      "loss": 0.5413,
+      "step": 331
+    },
+    {
+      "epoch": 2.1150159744408947,
+      "grad_norm": 2.6901578233118477,
+      "learning_rate": 2.3930766638769325e-06,
+      "loss": 0.3965,
+      "step": 332
+    },
+    {
+      "epoch": 2.121405750798722,
+      "grad_norm": 2.2464354768206967,
+      "learning_rate": 2.3613111881907273e-06,
+      "loss": 0.4243,
+      "step": 333
+    },
+    {
+      "epoch": 2.1277955271565494,
+      "grad_norm": 2.3073725414205857,
+      "learning_rate": 2.3296926463057396e-06,
+      "loss": 0.3954,
+      "step": 334
+    },
+    {
+      "epoch": 2.134185303514377,
+      "grad_norm": 2.317628398664997,
+      "learning_rate": 2.29822279888148e-06,
+      "loss": 0.4155,
+      "step": 335
+    },
+    {
+      "epoch": 2.1405750798722045,
+      "grad_norm": 1.9196452650507594,
+      "learning_rate": 2.2669033982974946e-06,
+      "loss": 0.3985,
+      "step": 336
+    },
+    {
+      "epoch": 2.146964856230032,
+      "grad_norm": 2.5155514594882864,
+      "learning_rate": 2.235736188555787e-06,
+      "loss": 0.3367,
+      "step": 337
+    },
+    {
+      "epoch": 2.1533546325878596,
+      "grad_norm": 2.2857947303762924,
+      "learning_rate": 2.2047229051837107e-06,
+      "loss": 0.4553,
+      "step": 338
+    },
+    {
+      "epoch": 2.159744408945687,
+      "grad_norm": 2.1456084476870556,
+      "learning_rate": 2.173865275137314e-06,
+      "loss": 0.4547,
+      "step": 339
+    },
+    {
+      "epoch": 2.1661341853035143,
+      "grad_norm": 2.240945856540863,
+      "learning_rate": 2.143165016705192e-06,
+      "loss": 0.4202,
+      "step": 340
+    },
+    {
+      "epoch": 2.1725239616613417,
+      "grad_norm": 1.9832713189435116,
+      "learning_rate": 2.1126238394127868e-06,
+      "loss": 0.4712,
+      "step": 341
+    },
+    {
+      "epoch": 2.1789137380191694,
+      "grad_norm": 1.9446053940500647,
+      "learning_rate": 2.082243443927212e-06,
+      "loss": 0.4358,
+      "step": 342
+    },
+    {
+      "epoch": 2.1853035143769968,
+      "grad_norm": 1.9350192214446225,
+      "learning_rate": 2.052025521962534e-06,
+      "loss": 0.5005,
+      "step": 343
+    },
+    {
+      "epoch": 2.191693290734824,
+      "grad_norm": 2.1868911275031033,
+      "learning_rate": 2.0219717561855857e-06,
+      "loss": 0.5543,
+      "step": 344
+    },
+    {
+      "epoch": 2.198083067092652,
+      "grad_norm": 2.221575157438669,
+      "learning_rate": 1.992083820122259e-06,
+      "loss": 0.3309,
+      "step": 345
+    },
+    {
+      "epoch": 2.2044728434504792,
+      "grad_norm": 2.144845801717813,
+      "learning_rate": 1.962363378064316e-06,
+      "loss": 0.3916,
+      "step": 346
+    },
+    {
+      "epoch": 2.2108626198083066,
+      "grad_norm": 2.11978548855504,
+      "learning_rate": 1.9328120849767198e-06,
+      "loss": 0.4164,
+      "step": 347
+    },
+    {
+      "epoch": 2.2172523961661343,
+      "grad_norm": 2.0385952191232155,
+      "learning_rate": 1.9034315864054682e-06,
+      "loss": 0.3998,
+      "step": 348
+    },
+    {
+      "epoch": 2.2236421725239617,
+      "grad_norm": 2.0754621683125207,
+      "learning_rate": 1.8742235183859747e-06,
+      "loss": 0.4357,
+      "step": 349
+    },
+    {
+      "epoch": 2.230031948881789,
+      "grad_norm": 2.0029050959618995,
+      "learning_rate": 1.8451895073519643e-06,
+      "loss": 0.3365,
+      "step": 350
+    },
+    {
+      "epoch": 2.236421725239617,
+      "grad_norm": 2.57807232565894,
+      "learning_rate": 1.8163311700448899e-06,
+      "loss": 0.3779,
+      "step": 351
+    },
+    {
+      "epoch": 2.242811501597444,
+      "grad_norm": 2.0445037247222464,
+      "learning_rate": 1.7876501134239316e-06,
+      "loss": 0.4552,
+      "step": 352
+    },
+    {
+      "epoch": 2.2492012779552715,
+      "grad_norm": 2.450940653052245,
+      "learning_rate": 1.7591479345764972e-06,
+      "loss": 0.4706,
+      "step": 353
+    },
+    {
+      "epoch": 2.255591054313099,
+      "grad_norm": 2.199833323100329,
+      "learning_rate": 1.7308262206292898e-06,
+      "loss": 0.4632,
+      "step": 354
+    },
+    {
+      "epoch": 2.2619808306709266,
+      "grad_norm": 2.653570092929433,
+      "learning_rate": 1.7026865486599375e-06,
+      "loss": 0.4422,
+      "step": 355
+    },
+    {
+      "epoch": 2.268370607028754,
+      "grad_norm": 2.3108018018174548,
+      "learning_rate": 1.6747304856091662e-06,
+      "loss": 0.4071,
+      "step": 356
+    },
+    {
+      "epoch": 2.2747603833865817,
+      "grad_norm": 2.2760801804733237,
+      "learning_rate": 1.6469595881935523e-06,
+      "loss": 0.4219,
+      "step": 357
+    },
+    {
+      "epoch": 2.281150159744409,
+      "grad_norm": 2.2663277836691185,
+      "learning_rate": 1.6193754028188363e-06,
+      "loss": 0.431,
+      "step": 358
+    },
+    {
+      "epoch": 2.2875399361022364,
+      "grad_norm": 2.1298943253664664,
+      "learning_rate": 1.591979465493806e-06,
+      "loss": 0.5151,
+      "step": 359
+    },
+    {
+      "epoch": 2.2939297124600637,
+      "grad_norm": 2.477752948045151,
+      "learning_rate": 1.5647733017447741e-06,
+      "loss": 0.5818,
+      "step": 360
+    },
+    {
+      "epoch": 2.3003194888178915,
+      "grad_norm": 2.219026126840176,
+      "learning_rate": 1.5377584265306222e-06,
+      "loss": 0.3346,
+      "step": 361
+    },
+    {
+      "epoch": 2.306709265175719,
+      "grad_norm": 2.0948183941547254,
+      "learning_rate": 1.510936344158448e-06,
+      "loss": 0.4208,
+      "step": 362
+    },
+    {
+      "epoch": 2.313099041533546,
+      "grad_norm": 2.4004015642072454,
+      "learning_rate": 1.484308548199796e-06,
+      "loss": 0.4439,
+      "step": 363
+    },
+    {
+      "epoch": 2.319488817891374,
+      "grad_norm": 2.0838816889076663,
+      "learning_rate": 1.4578765214074842e-06,
+      "loss": 0.3523,
+      "step": 364
+    },
+    {
+      "epoch": 2.3258785942492013,
+      "grad_norm": 2.429279030759271,
+      "learning_rate": 1.4316417356330441e-06,
+      "loss": 0.527,
+      "step": 365
+    },
+    {
+      "epoch": 2.3322683706070286,
+      "grad_norm": 2.2937409202550016,
+      "learning_rate": 1.4056056517447637e-06,
+      "loss": 0.4359,
+      "step": 366
+    },
+    {
+      "epoch": 2.3386581469648564,
+      "grad_norm": 2.335568744595556,
+      "learning_rate": 1.3797697195463278e-06,
+      "loss": 0.4426,
+      "step": 367
+    },
+    {
+      "epoch": 2.3450479233226837,
+      "grad_norm": 2.517760575107532,
+      "learning_rate": 1.3541353776961035e-06,
+      "loss": 0.4537,
+      "step": 368
+    },
+    {
+      "epoch": 2.351437699680511,
+      "grad_norm": 2.1836169125425493,
+      "learning_rate": 1.3287040536270135e-06,
+      "loss": 0.4155,
+      "step": 369
+    },
+    {
+      "epoch": 2.357827476038339,
+      "grad_norm": 2.206711961133346,
+      "learning_rate": 1.30347716346706e-06,
+      "loss": 0.5064,
+      "step": 370
+    },
+    {
+      "epoch": 2.364217252396166,
+      "grad_norm": 2.090805605064075,
+      "learning_rate": 1.2784561119604683e-06,
+      "loss": 0.3935,
+      "step": 371
+    },
+    {
+      "epoch": 2.3706070287539935,
+      "grad_norm": 2.245994580087271,
+      "learning_rate": 1.2536422923894565e-06,
+      "loss": 0.4418,
+      "step": 372
+    },
+    {
+      "epoch": 2.376996805111821,
+      "grad_norm": 1.7578744893397948,
+      "learning_rate": 1.2290370864966623e-06,
+      "loss": 0.3432,
+      "step": 373
+    },
+    {
+      "epoch": 2.3833865814696487,
+      "grad_norm": 2.0933614656607897,
+      "learning_rate": 1.2046418644081904e-06,
+      "loss": 0.3342,
+      "step": 374
+    },
+    {
+      "epoch": 2.389776357827476,
+      "grad_norm": 2.0808115537211695,
+      "learning_rate": 1.1804579845573288e-06,
+      "loss": 0.5301,
+      "step": 375
+    },
+    {
+      "epoch": 2.3961661341853033,
+      "grad_norm": 1.9927356398390765,
+      "learning_rate": 1.156486793608899e-06,
+      "loss": 0.4528,
+      "step": 376
+    },
+    {
+      "epoch": 2.402555910543131,
+      "grad_norm": 1.9488088695501393,
+      "learning_rate": 1.1327296263842653e-06,
+      "loss": 0.5109,
+      "step": 377
+    },
+    {
+      "epoch": 2.4089456869009584,
+      "grad_norm": 2.363258143968508,
+      "learning_rate": 1.1091878057870137e-06,
+      "loss": 0.4615,
+      "step": 378
+    },
+    {
+      "epoch": 2.415335463258786,
+      "grad_norm": 2.214690728832497,
+      "learning_rate": 1.0858626427292796e-06,
+      "loss": 0.3428,
+      "step": 379
+    },
+    {
+      "epoch": 2.4217252396166136,
+      "grad_norm": 2.4024816832009157,
+      "learning_rate": 1.0627554360587533e-06,
+      "loss": 0.4131,
+      "step": 380
+    },
+    {
+      "epoch": 2.428115015974441,
+      "grad_norm": 1.7023192512400525,
+      "learning_rate": 1.0398674724863584e-06,
+      "loss": 0.3013,
+      "step": 381
+    },
+    {
+      "epoch": 2.4345047923322682,
+      "grad_norm": 2.5554590548546408,
+      "learning_rate": 1.0172000265145938e-06,
+      "loss": 0.429,
+      "step": 382
+    },
+    {
+      "epoch": 2.440894568690096,
+      "grad_norm": 2.4406043120968146,
+      "learning_rate": 9.947543603665711e-07,
+      "loss": 0.4446,
+      "step": 383
+    },
+    {
+      "epoch": 2.4472843450479234,
+      "grad_norm": 2.355132900119845,
+      "learning_rate": 9.72531723915726e-07,
+      "loss": 0.4686,
+      "step": 384
+    },
+    {
+      "epoch": 2.4536741214057507,
+      "grad_norm": 2.009745480448959,
+      "learning_rate": 9.505333546162171e-07,
+      "loss": 0.4943,
+      "step": 385
+    },
+    {
+      "epoch": 2.460063897763578,
+      "grad_norm": 2.319627194265683,
+      "learning_rate": 9.287604774340236e-07,
+      "loss": 0.4987,
+      "step": 386
+    },
+    {
+      "epoch": 2.466453674121406,
+      "grad_norm": 2.434761883812788,
+      "learning_rate": 9.07214304778729e-07,
+      "loss": 0.4593,
+      "step": 387
+    },
+    {
+      "epoch": 2.472843450479233,
+      "grad_norm": 1.95171692263905,
+      "learning_rate": 8.858960364360142e-07,
+      "loss": 0.3701,
+      "step": 388
+    },
+    {
+      "epoch": 2.479233226837061,
+      "grad_norm": 2.2046486182188065,
+      "learning_rate": 8.648068595008458e-07,
+      "loss": 0.4211,
+      "step": 389
+    },
+    {
+      "epoch": 2.4856230031948883,
+      "grad_norm": 2.6042374898410787,
+      "learning_rate": 8.439479483113683e-07,
+      "loss": 0.4559,
+      "step": 390
+    },
+    {
+      "epoch": 2.4920127795527156,
+      "grad_norm": 2.0963323630173014,
+      "learning_rate": 8.233204643835235e-07,
+      "loss": 0.3564,
+      "step": 391
+    },
+    {
+      "epoch": 2.498402555910543,
+      "grad_norm": 2.4092444760160077,
+      "learning_rate": 8.029255563463589e-07,
+      "loss": 0.4075,
+      "step": 392
+    },
+    {
+      "epoch": 2.5047923322683707,
+      "grad_norm": 2.4649195547395064,
+      "learning_rate": 7.827643598780748e-07,
+      "loss": 0.4616,
+      "step": 393
+    },
+    {
+      "epoch": 2.511182108626198,
+      "grad_norm": 2.0786671044545595,
+      "learning_rate": 7.628379976427868e-07,
+      "loss": 0.4258,
+      "step": 394
+    },
+    {
+      "epoch": 2.5175718849840254,
+      "grad_norm": 2.346266411020929,
+      "learning_rate": 7.431475792280018e-07,
+      "loss": 0.4013,
+      "step": 395
+    },
+    {
+      "epoch": 2.523961661341853,
+      "grad_norm": 2.0883565799461197,
+      "learning_rate": 7.23694201082843e-07,
+      "loss": 0.4058,
+      "step": 396
+    },
+    {
+      "epoch": 2.5303514376996805,
+      "grad_norm": 1.998904610700326,
+      "learning_rate": 7.044789464569817e-07,
+      "loss": 0.3783,
+      "step": 397
+    },
+    {
+      "epoch": 2.536741214057508,
+      "grad_norm": 2.230046009669308,
+      "learning_rate": 6.855028853403295e-07,
+      "loss": 0.4122,
+      "step": 398
+    },
+    {
+      "epoch": 2.543130990415335,
+      "grad_norm": 1.9887522365789791,
+      "learning_rate": 6.667670744034498e-07,
+      "loss": 0.4234,
+      "step": 399
+    },
+    {
+      "epoch": 2.549520766773163,
+      "grad_norm": 2.577199407242772,
+      "learning_rate": 6.482725569387171e-07,
+      "loss": 0.3888,
+      "step": 400
+    },
+    {
+      "epoch": 2.5559105431309903,
+      "grad_norm": 2.402143623444451,
+      "learning_rate": 6.300203628022272e-07,
+      "loss": 0.4289,
+      "step": 401
+    },
+    {
+      "epoch": 2.562300319488818,
+      "grad_norm": 2.1944070015219888,
+      "learning_rate": 6.120115083564432e-07,
+      "loss": 0.4336,
+      "step": 402
+    },
+    {
+      "epoch": 2.5686900958466454,
+      "grad_norm": 2.0589970878526844,
+      "learning_rate": 5.942469964136055e-07,
+      "loss": 0.3676,
+      "step": 403
+    },
+    {
+      "epoch": 2.5750798722044728,
+      "grad_norm": 2.502670968119668,
+      "learning_rate": 5.767278161798912e-07,
+      "loss": 0.3946,
+      "step": 404
+    },
+    {
+      "epoch": 2.5814696485623,
+      "grad_norm": 2.234483396791303,
+      "learning_rate": 5.594549432003244e-07,
+      "loss": 0.4159,
+      "step": 405
+    },
+    {
+      "epoch": 2.587859424920128,
+      "grad_norm": 2.508675316184731,
+      "learning_rate": 5.42429339304461e-07,
+      "loss": 0.3766,
+      "step": 406
+    },
+    {
+      "epoch": 2.594249201277955,
+      "grad_norm": 2.2388652923811097,
+      "learning_rate": 5.256519525528254e-07,
+      "loss": 0.3741,
+      "step": 407
+    },
+    {
+      "epoch": 2.600638977635783,
+      "grad_norm": 2.1367021327758926,
+      "learning_rate": 5.091237171841173e-07,
+      "loss": 0.509,
+      "step": 408
+    },
+    {
+      "epoch": 2.6070287539936103,
+      "grad_norm": 2.1883188048010505,
+      "learning_rate": 4.92845553563196e-07,
+      "loss": 0.3475,
+      "step": 409
+    },
+    {
+      "epoch": 2.6134185303514377,
+      "grad_norm": 2.6958069924361903,
+      "learning_rate": 4.768183681298211e-07,
+      "loss": 0.5423,
+      "step": 410
+    },
+    {
+      "epoch": 2.619808306709265,
+      "grad_norm": 2.3649443474581617,
+      "learning_rate": 4.6104305334818577e-07,
+      "loss": 0.4518,
+      "step": 411
+    },
+    {
+      "epoch": 2.626198083067093,
+      "grad_norm": 2.7579147997887103,
+      "learning_rate": 4.455204876572172e-07,
+      "loss": 0.3223,
+      "step": 412
+    },
+    {
+      "epoch": 2.63258785942492,
+      "grad_norm": 2.322240849008944,
+      "learning_rate": 4.3025153542165744e-07,
+      "loss": 0.3889,
+      "step": 413
+    },
+    {
+      "epoch": 2.6389776357827475,
+      "grad_norm": 2.2585577552913327,
+      "learning_rate": 4.1523704688394176e-07,
+      "loss": 0.4524,
+      "step": 414
+    },
+    {
+      "epoch": 2.6453674121405752,
+      "grad_norm": 2.21502448661817,
+      "learning_rate": 4.0047785811684116e-07,
+      "loss": 0.4312,
+      "step": 415
+    },
+    {
+      "epoch": 2.6517571884984026,
+      "grad_norm": 2.355337885891376,
+      "learning_rate": 3.8597479097691626e-07,
+      "loss": 0.4265,
+      "step": 416
+    },
+    {
+      "epoch": 2.65814696485623,
+      "grad_norm": 2.2244869149661306,
+      "learning_rate": 3.717286530587483e-07,
+      "loss": 0.5291,
+      "step": 417
+    },
+    {
+      "epoch": 2.6645367412140573,
+      "grad_norm": 2.3569227035622013,
+      "learning_rate": 3.577402376499672e-07,
+      "loss": 0.4332,
+      "step": 418
+    },
+    {
+      "epoch": 2.670926517571885,
+      "grad_norm": 2.133768443895229,
+      "learning_rate": 3.440103236870823e-07,
+      "loss": 0.4609,
+      "step": 419
+    },
+    {
+      "epoch": 2.6773162939297124,
+      "grad_norm": 2.3164553548368323,
+      "learning_rate": 3.3053967571210375e-07,
+      "loss": 0.479,
+      "step": 420
+    },
+    {
+      "epoch": 2.68370607028754,
+      "grad_norm": 2.3702331769156735,
+      "learning_rate": 3.1732904382996975e-07,
+      "loss": 0.4552,
+      "step": 421
+    },
+    {
+      "epoch": 2.6900958466453675,
+      "grad_norm": 2.1694566517033724,
+      "learning_rate": 3.04379163666782e-07,
+      "loss": 0.3982,
+      "step": 422
+    },
+    {
+      "epoch": 2.696485623003195,
+      "grad_norm": 2.507080592863833,
+      "learning_rate": 2.916907563288357e-07,
+      "loss": 0.495,
+      "step": 423
+    },
+    {
+      "epoch": 2.702875399361022,
+      "grad_norm": 1.9030589659154709,
+      "learning_rate": 2.792645283624712e-07,
+      "loss": 0.355,
+      "step": 424
+    },
+    {
+      "epoch": 2.70926517571885,
+      "grad_norm": 2.486303979910709,
+      "learning_rate": 2.671011717147276e-07,
+      "loss": 0.4208,
+      "step": 425
+    },
+    {
+      "epoch": 2.7156549520766773,
+      "grad_norm": 2.6926299813084595,
+      "learning_rate": 2.5520136369481194e-07,
+      "loss": 0.4193,
+      "step": 426
+    },
+    {
+      "epoch": 2.722044728434505,
+      "grad_norm": 2.4101859848112404,
+      "learning_rate": 2.4356576693638555e-07,
+      "loss": 0.4106,
+      "step": 427
+    },
+    {
+      "epoch": 2.7284345047923324,
+      "grad_norm": 2.1092564449268303,
+      "learning_rate": 2.3219502936066228e-07,
+      "loss": 0.4649,
+      "step": 428
+    },
+    {
+      "epoch": 2.7348242811501597,
+      "grad_norm": 2.409939316965895,
+      "learning_rate": 2.210897841403331e-07,
+      "loss": 0.4726,
+      "step": 429
+    },
+    {
+      "epoch": 2.741214057507987,
+      "grad_norm": 2.9923889890507893,
+      "learning_rate": 2.1025064966430697e-07,
+      "loss": 0.4357,
+      "step": 430
+    },
+    {
+      "epoch": 2.747603833865815,
+      "grad_norm": 1.9116249904508271,
+      "learning_rate": 1.9967822950327453e-07,
+      "loss": 0.3906,
+      "step": 431
+    },
+    {
+      "epoch": 2.753993610223642,
+      "grad_norm": 2.363072327163832,
+      "learning_rate": 1.8937311237610168e-07,
+      "loss": 0.4384,
+      "step": 432
+    },
+    {
+      "epoch": 2.7603833865814695,
+      "grad_norm": 2.111336234633767,
+      "learning_rate": 1.793358721170435e-07,
+      "loss": 0.3841,
+      "step": 433
+    },
+    {
+      "epoch": 2.7667731629392973,
+      "grad_norm": 2.4061012547366913,
+      "learning_rate": 1.6956706764379438e-07,
+      "loss": 0.413,
+      "step": 434
+    },
+    {
+      "epoch": 2.7731629392971247,
+      "grad_norm": 2.3834000028849007,
+      "learning_rate": 1.6006724292636166e-07,
+      "loss": 0.4789,
+      "step": 435
+    },
+    {
+      "epoch": 2.779552715654952,
+      "grad_norm": 2.2642059928473066,
+      "learning_rate": 1.508369269567783e-07,
+      "loss": 0.4806,
+      "step": 436
+    },
+    {
+      "epoch": 2.7859424920127793,
+      "grad_norm": 2.1091265211005017,
+      "learning_rate": 1.418766337196431e-07,
+      "loss": 0.3913,
+      "step": 437
+    },
+    {
+      "epoch": 2.792332268370607,
+      "grad_norm": 2.0974258225125464,
+      "learning_rate": 1.3318686216350241e-07,
+      "loss": 0.401,
+      "step": 438
+    },
+    {
+      "epoch": 2.7987220447284344,
+      "grad_norm": 2.725505526625615,
+      "learning_rate": 1.2476809617306408e-07,
+      "loss": 0.6157,
+      "step": 439
+    },
+    {
+      "epoch": 2.8051118210862622,
+      "grad_norm": 2.3394680432484787,
+      "learning_rate": 1.166208045422551e-07,
+      "loss": 0.4105,
+      "step": 440
+    },
+    {
+      "epoch": 2.8115015974440896,
+      "grad_norm": 2.4705712014808996,
+      "learning_rate": 1.0874544094811424e-07,
+      "loss": 0.3823,
+      "step": 441
+    },
+    {
+      "epoch": 2.817891373801917,
+      "grad_norm": 2.0392789957555753,
+      "learning_rate": 1.0114244392553318e-07,
+      "loss": 0.44,
+      "step": 442
+    },
+    {
+      "epoch": 2.8242811501597442,
+      "grad_norm": 2.348950535667093,
+      "learning_rate": 9.381223684283291e-08,
+      "loss": 0.4619,
+      "step": 443
+    },
+    {
+      "epoch": 2.830670926517572,
+      "grad_norm": 2.1136302016617683,
+      "learning_rate": 8.675522787819023e-08,
+      "loss": 0.371,
+      "step": 444
+    },
+    {
+      "epoch": 2.8370607028753994,
+      "grad_norm": 2.3859707399616203,
+      "learning_rate": 7.997180999691101e-08,
+      "loss": 0.4676,
+      "step": 445
+    },
+    {
+      "epoch": 2.8434504792332267,
+      "grad_norm": 2.3594281771562993,
+      "learning_rate": 7.346236092954318e-08,
+      "loss": 0.496,
+      "step": 446
+    },
+    {
+      "epoch": 2.8498402555910545,
+      "grad_norm": 2.289174122732648,
+      "learning_rate": 6.722724315084805e-08,
+      "loss": 0.4047,
+      "step": 447
+    },
+    {
+      "epoch": 2.856230031948882,
+      "grad_norm": 2.378937591165076,
+      "learning_rate": 6.12668038596137e-08,
+      "loss": 0.4415,
+      "step": 448
+    },
+    {
+      "epoch": 2.862619808306709,
+      "grad_norm": 2.0870909731738823,
+      "learning_rate": 5.5581374959320366e-08,
+      "loss": 0.4469,
+      "step": 449
+    },
+    {
+      "epoch": 2.8690095846645365,
+      "grad_norm": 2.2446199520479344,
+      "learning_rate": 5.017127303966085e-08,
+      "loss": 0.4002,
+      "step": 450
+    },
+    {
+      "epoch": 2.8753993610223643,
+      "grad_norm": 2.2471332517670275,
+      "learning_rate": 4.50367993589107e-08,
+      "loss": 0.4317,
+      "step": 451
+    },
+    {
+      "epoch": 2.8817891373801916,
+      "grad_norm": 2.0864772616550673,
+      "learning_rate": 4.0178239827151077e-08,
+      "loss": 0.3966,
+      "step": 452
+    },
+    {
+      "epoch": 2.8881789137380194,
+      "grad_norm": 2.2798143023681794,
+      "learning_rate": 3.559586499035206e-08,
+      "loss": 0.3958,
+      "step": 453
+    },
+    {
+      "epoch": 2.8945686900958467,
+      "grad_norm": 2.1208404230304145,
+      "learning_rate": 3.128993001530245e-08,
+      "loss": 0.4001,
+      "step": 454
+    },
+    {
+      "epoch": 2.900958466453674,
+      "grad_norm": 2.1876514381127716,
+      "learning_rate": 2.7260674675404498e-08,
+      "loss": 0.3963,
+      "step": 455
+    },
+    {
+      "epoch": 2.9073482428115014,
+      "grad_norm": 2.2369393890230924,
+      "learning_rate": 2.3508323337321225e-08,
+      "loss": 0.412,
+      "step": 456
+    },
+    {
+      "epoch": 2.913738019169329,
+      "grad_norm": 1.9309934320730493,
+      "learning_rate": 2.0033084948483104e-08,
+      "loss": 0.3299,
+      "step": 457
+    },
+    {
+      "epoch": 2.9201277955271565,
+      "grad_norm": 2.265789089300665,
+      "learning_rate": 1.6835153025451246e-08,
+      "loss": 0.4279,
+      "step": 458
+    },
+    {
+      "epoch": 2.9265175718849843,
+      "grad_norm": 2.1236508228489885,
+      "learning_rate": 1.3914705643143788e-08,
+      "loss": 0.4297,
+      "step": 459
+    },
+    {
+      "epoch": 2.9329073482428116,
+      "grad_norm": 2.5202420167871775,
+      "learning_rate": 1.1271905424918294e-08,
+      "loss": 0.4684,
+      "step": 460
+    },
+    {
+      "epoch": 2.939297124600639,
+      "grad_norm": 2.223482125500666,
+      "learning_rate": 8.906899533517866e-09,
+      "loss": 0.4636,
+      "step": 461
+    },
+    {
+      "epoch": 2.9456869009584663,
+      "grad_norm": 1.9511784007469755,
+      "learning_rate": 6.819819662874372e-09,
+      "loss": 0.3447,
+      "step": 462
+    },
+    {
+      "epoch": 2.952076677316294,
+      "grad_norm": 2.484026001564462,
+      "learning_rate": 5.0107820307770945e-09,
+      "loss": 0.5442,
+      "step": 463
+    },
+    {
+      "epoch": 2.9584664536741214,
+      "grad_norm": 2.703733619110161,
+      "learning_rate": 3.4798873723984604e-09,
+      "loss": 0.4034,
+      "step": 464
+    },
+    {
+      "epoch": 2.9648562300319488,
+      "grad_norm": 2.1404889485604897,
+      "learning_rate": 2.2272209346885233e-09,
+      "loss": 0.4759,
+      "step": 465
+    },
+    {
+      "epoch": 2.9712460063897765,
+      "grad_norm": 2.0827067956874137,
+      "learning_rate": 1.2528524716259872e-09,
+      "loss": 0.4849,
+      "step": 466
+    },
+    {
+      "epoch": 2.977635782747604,
+      "grad_norm": 2.339494396070975,
+      "learning_rate": 5.568362403318706e-10,
+      "loss": 0.4075,
+      "step": 467
+    },
+    {
+      "epoch": 2.984025559105431,
+      "grad_norm": 2.2851445627861593,
+      "learning_rate": 1.3921099805302985e-10,
+      "loss": 0.3724,
+      "step": 468
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 468,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5655918280704.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9d2be5da7ffab4d71b98d14ef090756b9ff74621d72aee4e8cb58a912a6fc1dd
+size 8017